JERUSALEM STUDIES IN ARABIC AND ISLAM 36(2009)¡gyi, K - A prophet like Jesus.pdf132 Krisztina Szil agyi Muh.ammad’s death according to the Christians ... a Christian Arab belonging

The Institute of Asian and African StudiesThe Max Schloessinger Memorial Foundation

Offprint from

JERUSALEM STUDIES INARABIC AND ISLAM

36(2009)

Krisztina Szilagyi

A prophet like Jesus? Christians andMuslims debating Muh. ammad’s death

THE HEBREW UNIVERSITY OF JERUSALEMTHE FACULTY OF HUMANITIES

JSAI 36 (2009)

A PROPHET LIKE JESUS? CHRISTIANS ANDMUSLIMS DEBATING MUH. AMMAD’S DEATH∗

Krisztina SzilagyiPrinceton University

Scholars commonly accept that medieval Christian polemicists basedmuch of their representation of Muh.ammad’s life on ignorance and mis-understanding, even willful distortion of the Muslim tradition. This hasalso become the standard interpretation of the legend of Muh.ammad’sdeath that circulated among the Christians of the Islamic world. Thispolemical story recounts Muh.ammad’s death and its immediate after-math. It claims that Muh.ammad foretold that he would be resurrectedthree days after his death, yet while his followers delayed the burial inanticipation of his resurrection, his body started to exhibit signs of de-cay. As a scholar of medieval Latin Christendom asserts, in this legend“Muh.ammad’s death is described in a manner that has nothing to dowith Muslim tradition.”1

A careful examination of the Muslim tradition, however, suggestsotherwise. The purpose of this paper is to show that, although the fullstory fundamentally differs from the classical Islamic narrative of theProphet’s death, each of its motifs save one appears in Muslim literature.Often they occur independently but sometimes also in combination witheach other, suggesting that Christians borrowed most of the narrativedirectly from the Islamic tradition. I therefore argue that, rather thanbeing ignorant, some Christians had sufficiently deep knowledge of theMuslim tradition to make a sophisticated selection of h. adıths that weresuitable for their polemical purposes. The first part of the paper surveysand analyzes the surviving versions of the Christian legend, while thenext two examine their Muslim sources. The fourth part attempts totrace the interrelationship of these narratives against the background ofthe polemical milieu of the eighth century.

∗I am grateful to Professor Michael Cook, Professor Patricia Crone, and Dr. Ed-uard Iricinschi for their comments on this paper.

1Tolan, Saracens, p. 92. For more similar opinions about the legend, see below,p. 138, n. 25.

131

132 Krisztina Szilagyi

Muh. ammad’s death according to the Christians

The oldest extant Christian texts containing the Christian legend ofMuh.ammad’s death date from the ninth century. Their authors livedin distant parts of the Islamic world, belonged to different Christiancommunities and spoke different languages. In the East, the narrativeappears in the Syriac recensions of the Christian Bah. ıra legend, a pop-ular Christian counterhistory of the rise of Islam.2 The oldest versionof the Christian Bah. ıra legend was most probably written by a West-Syrian monk in Iraq in the 810s; the story of Muh.ammad’s death oc-curs in two recensions, a West-Syrian and an East-Syrian. They shouldprobably be dated to the middle of the ninth and to the tenth century,respectively.3 Another text, the Apology of al-Kindı, an Arabic polemi-cal treatise against Islam, has a detailed biography of Muh.ammad whichincludes the legend.4 Its author was probably an East-Syrian courtier ofal-Ma↩mun (813–33), a Christian Arab belonging to the tribe of Kinda,and it seems that he wrote the Apology in the 820s in Baghdad.5 A

2On the Christian Bah. ıra legend, see Roggema, The legend of Sergius Bah. ıra, pp.11–208.

3For the two versions, see ibid., pp. 302–303 (East-Syrian, Syriac and Englishtext), pp. 334–335 (West-Syrian, Syriac and English). The two Arabic recensionsomit the episode. The dating of the Christian Bah. ıra legend to the reign of al-Ma↩mun (r. 813–33) is now generally accepted; see ibid., pp. 86–87. For an attemptto date the individual recensions more precisely, see Szilagyi, “Muh. ammad and themonk,” p. 191. The story of Muh. ammad’s death in the East-Syrian recension doesnot strictly belong to the Christian Bah. ıra legend, since it is found in an appendixwith a separate heading; therefore, its date is uncertain (see ibid., pp. 177–178, 192).Its proximity to the version in the West-Syrian recension (see below, pp. 133–134)nevertheless justifies the discussion of these two together.

4About the Apology of al-Kindı, see Samir, “Apologie d’al-Kindı;” Koningsveld,“The Apology of Al-Kindı;” Landron, Attitudes nestoriennes, pp. 78–88. There isno critical edition of the text; the most readily available printing, based on twomanuscripts, is Tien, Risala; another one, based on four manuscripts, is Tartar,Dialogue islamo-chretien. For the Apology’s discussion of Muh. ammad’s death, seeTien, Risala, pp. 109–110; and Tartar, Dialogue islamo-chretien, pp. 92–93.

The Apology was translated into Latin in 1142. This translation sometimes reflectsan earlier stage of the text than the published Arabic versions that are based onmanuscripts copied in the seventeenth century or later. The Latin rendering of thestory of Muh. ammad’s death is, however, very similar to the Arabic text. For therelationship of the Latin and the Arabic texts, see Koningsveld, “The Apology ofal-Kindı,” pp. 70–75; Samir, “Apologie d’al-Kindı,” pp. 48–74; for the most recentcritical edition of the Latin translation, see Gonzalez Munoz, Exposicion y refutaciondel Islam.

5The dating and the authorship of the Apology are controversial (for a recentoverview of the arguments for the various positions, see Samir, “Apologie d’al-Kindı,”

Christians and Muslims debating Muh. ammad’s death 133

lesser-known text that contains the legend is the Qashun document.The Qashun document, although extant only in Armenian translationand first attested only in the thirteenth century, appears to preserve aChristian account on Muh.ammad’s life and the rise of Islam from thelate eighth- or early ninth century Iraq. Its original language was prob-ably Syriac or Arabic.6 The legend was known in Spain too. Eulogiusof Cordoba (d. 859), learned priest, supporter of the Cordoban martyrs,and eventually one of them, included a brief Latin life of Muh.ammad, Is-toria de Mahomet, in his Liber apologeticus martyrum. Eulogius remarksthat he found the Istoria in the monastery of Leyre beside Pamplona,during his travels in Northern Spain in 849–50. The Istoria ends withthe Christian legend of Muh.ammad’s death, its longest extant version.Finally, a note on Muh.ammad’s life, Adnotatio de Mammet that alsomentions the story, appears in the letter of John of Seville, addressed tothe Cordoban Paul Albar (d. ca. 861). John probably wrote the letterbefore 851.7

The legend as given in the West-Syrian recension of the ChristianBah. ıra legend is a typical example of the shorter versions:8

He [Ka↪b al-Ah.bar] said to them, “. . . There will be a signto you: When Muh.ammad dies he will ascend to heaven like↪Isa, son of Maryam, and will be resurrected after three days.”

pp. 33–41). The text refers to the revolt of Babak al-Khurramı (from 816 or 817 to837 AD) as contemporary, and states that little more than 200 years (nayyif wa-mi ↩ata sana) passed since the time of Muh. ammad (200 AH = 815–6 CE) (Tien,Risala, pp. 76, 97; Tartar, Dialogue islamo-chretien, pp. 68, 84), but a tenth centurydating has been proposed by some scholars on the basis of what they see as thereliance of al-Kindı on late ninth- or early tenth-century works. It is not possible todiscuss this issue here in detail, but al-Kindı’s references to the early ninth century,the Islamic sources he used and the Muslim authors he mentioned make the Apology,in my opinion, fit the early ninth century much better than the tenth.

6The Qashun document has been translated into English in Thomson,“Muh. ammad,” pp. 846–853; the story of Muh. ammad’s death is told on p. 850, to beread with notes y–z. On the dating and provenance of this text, see the Appendix,pp. 159–162.

7On Paulus Alvarus, see Colbert, The martyrs of Cordoba, pp. 148–166, 305–332;on the Adnotatio, see ibid., pp. 156–157. About Eulogius, see Colbert, The martyrsof Cordoba, pp. 174–222; about the Istoria, see ibid., pp. 334–339, and Wolf, “Theearliest Latin lives,” pp. 89–100. The Latin text of the Istoria has been edited inDıaz y Dıaz, “Los textos antimahometanos,” pp. 157–159; Gil, Corpus scriptorummuzarabicorum, vol. 2, pp. 483–486; for English translations, see Wolf, “The earliestLatin lives,” pp. 97–99, and Colbert, The martyrs of Cordoba, pp. 336–339. For theLatin text of the Adnotatio, see Dıaz y Dıaz, “Textos antimahometanos,” p. 153; Gil,Corpus scriptorum muzarabicorum, vol. 1, pp. 200–201; for an English translationsee Hoyland, Seeing Islam, p. 513.

8Roggema, The legend of Sergius Bah. ıra, p. 335 (modified).


It happened that when Muh. ammad died his kinsmen assem-bled, embalmed him, and laid him in a house with greatreverence. They sealed the door on him to see what wouldbecome of him.

Three days later they opened the door, but nobody couldenter the house because of the stink of Muh.ammad’s body.No one needs to investigate what happened to it.

The version of the story in the West-Syrian recension of the ChristianBah. ıra legend attributes the prophecy of Muh.ammad’s resurrection fromthe dead to Ka↪b al-Ah. bar. No other version does so. The East-Syrianrecension of the Christian Bah. ıra legend leaves the prophecy’s origin un-specified, and the rest ascribe it to Muh.ammad himself. Both the Syriacrecensions of the Christian Bah. ıra legend and the Apology of al-Kindıspeak of an expectation of ascension;9 the other Christian texts men-tion resurrection instead. The version of the Istoria, the most elaborateone, includes the unique detail that the Muslims expected “the angelGabriel” (later in the text “angels”) to come and revive Muh.ammad.The anticipated resurrection is explicitly compared to Christ’s story inall versions, with the exception of the two Latin texts. His followerssecure Muh. ammad’s corpse in a locked house in both Syriac recensionsof the Christian Bah. ıra legend, and they lay it out in his garden inthe Qashun document. The remaining versions do not specify a place.Guards watch over the body according to the two Latin versions; “dis-ciples” do the same according to the Qashun document. In both Syriacrecensions of the Christian Bah. ıra legend and the Apology of al-Kindı,the Muslims realize the futility of their expectation on the third day whenthey notice that the decomposition of the corpse had already started. Inthe other texts, their hopes come to an end when dogs devour the decay-ing cadaver. As we read it in the Istoria, “. . . dogs followed his stenchand devoured his flank. Learning of the deed, they surrendered the restof his body to the soil. And in vindication of this injury, they ordereddogs to be slaughtered every year. . . ”10 A yearly massacre of dogs wasinstituted also according to the Qashun document; it is, the text claims,observed “up to the present day.”

This comparison of motifs shows that the five early Christian versionsof the legend fall into two subgroups. The first includes the versions thatappear in the two Syriac recensions of the Christian Bah. ıra legend andin the Apology of al-Kindı, the former two being more closely related to

9The resurrection, since it is referred to only after the ascension, appears to besecondary in the West-Syrian recension.

10Wolf, “The earliest Latin lives,” p. 99.


each other than to the latter. The second comprises the two Latin textsand the Qashun document. The former two of this group are also moresimilar to each other than to the latter.

The wide dissemination of the Christian story by the middle of theninth century indicates an older origin. The texts of the second subgroupshow roots earlier than the ninth century. But there is reason to believethat the legend circulated already in the first half of the eighth century.The Qashun document, although attested only in later Armenian ver-sions, is likely to go back to a Syriac or Arabic text on the origins ofIslam compiled in Iraq during the first decade of the ninth century atthe latest. Its author probably relied on several written and oral sourcesof various ages and provenances, and one of his written sources mighthave been produced as early as the seventh century.11 It is not possible,however, to establish the origin of the individual parts of the text withcertainty. The two Latin texts, the Istoria and the Adnotatio, are morehelpful in this respect. Their comparison shows that both drew on thesame Latin source written in Spain. Their Latin source, in turn, is prob-ably based on an Eastern one. In view of the remarks of the Istoria andthe Adnotatio related to Byzantium, the Eastern source seems to havebeen of Melkite provenance. Since the Istoria refers to Damascus as thecapital of the Arabs, either its Spanish Latin source or the Melkite sourceof the latter was written before the end of the Umayyad caliphate. Also,the Spanish Latin source dates the appearance of Muh.ammad to theseventh year of the reign of Heraclius (610–41), similarly to the Hispanicchronicles of 741 and 754, which again points to its origin in the first halfof the eighth century.12 We can thus conclude that the Christian storyof Muh. ammad’s death was, in all probability, known among Christiansin the Near East during the first half of the eighth century.

In addition, Muslim authors occasionally quote a brief reference tothe story from an Egyptian Christian in their discussions of the punish-ment of non-Muslims who commit the crime of slandering the Prophet.According to the earliest extant citation, in the Shifa↩ of al-Qad. ı ↪Iyad. ,13

11See Appendix, pp. 159–162.12See the discussions about the interrelationship and the origin of the texts in

Colbert, The martyrs of Cordoba, pp. 156–157; Franke, Die freiwilligen Martyrervon Cordova, pp. 38–47; Dıaz y Dıaz, “Los textos antimahometanos,” pp. 165–168(appendix by I.B. Ceinos); Daniel, Arabs and mediaeval Europe, pp. 39–48; Hoyland,Seeing Islam, pp. 514–515.

13Al-Qad. ı ↪Iyad. , Shifa↩, vol. 2, pp. 1037–1039, quotation from pp. 1037–1038 (qalabnu ’l-qasimi sa↩alna malikan ↪an nas.raniyyin bi-mis.ra shuhida ↪alayhi annahu qalamiskınun muh. ammadun yukhbirukum annahu fı ’l-jannati maluhu [or ma lahu] lamyanfa↪ nafsahu idh kanati ’l-kilabu ta↩kulu saqayhi law qataluhu ’starah. a minhu ’l-nas qala malikun ara an tud. raba ’unuquhu). Safran, “Identity and differentiation,”p. 589 cites the incident.


Ibn al-Qasim said, “We asked Malik [b. Anas] about a Chris-tian in Egypt against whom there was testimony that he said,‘Poor Muh.ammad! He is telling you that he is in Paradise[but] his wealth did not benefit his soul when dogs were eating(ta↩kulu) his legs. Had he been killed, the people would havefound rest from him.’ Malik answered, ‘I think he should beexecuted.’”

In response to further inquiry from Ibn al-Qasim, whether the corpse ofthe executed Copt should be burned, Malik agreed that it should, andthe sentence, al-Qad. ı ↪Iyad. continues to quote his source, was carriedout.

Several Malikı works cite the story in full, and sometimes they nametheir source as the Shifa↩. Many curtail it, giving only the Copt’s wordsand Malik’s decision, in a long series of non-Muslim slanders of theProphet and their recommended punishment, collected for the edifica-tion of future generations. With two exceptions, their authors displayno knowledge of the incident apart from what could be gleaned fromthe Shifa↩.14 The first of the two exceptions, al-Wansharısı perhapsused a different source, but gives no further indication of the partici-pants or information about the circumstances.15 The second, Ibn H. ajaral-↪Asqalanı gives a better version of the Copt’s outburst than al-Qad. ı↪Iyad. , and identifies his source as al-H. arith b. Miskın, a ninth-centuryMalikı qad. ı of Egypt (d. 250/864) who in turn quotes it from Ibn al-Qasim, his teacher. ↪Abd al-Rah.man b. al-Qasim (d. 191/806) was themost prominent disciple of Malik b. Anas (d. 179/795), studied with himfor twenty years in Medina, then settled in Egypt and died there. IbnH. ajar furthermore identifies the qad. ı who carried out the punishmentas al-Mufad.d. al b. Fad. ala (d. 181/797), a contemporary of Malik.16 It ispossible that all these three quotations go back to a single earlier source,but the reference to several contemporary locals in the last version ascer-tains that the incident indeed took place in the second half of the eighthcentury in Egypt. It thus constitutes the only datable attestation of thestory before the ninth century, when it first appears in Christian sources,and the only one in Egypt.17 A Malikı judge would hardly have inventedthis lurid story. At the same time, the similarity of the quotations of

14See, for example, Qarafı, Dhakhıra, vol. 12, p. 20; S. alih. ı, Subul al-huda, vol. 12,p. 34; H. ashiyat al-Dasuqı, vol. 2, p. 205.

15Wansharısı, Mi ↪yar, vol. 2, p. 345.16Ibn H. ajar, Raf ↪ al-is.r, p. 440 (miskınun muh. ammadun yaqulu lakum innakum

fı ’l-jannati a-huwa al-ana fı ’l-jannati miskınun fa-maluhu la yanfa↪u nafsahu idhkanati ’l-kilabu ta↩kulu saqayhi law kana uh. riqa bi-’l-nari ’starah. a minhu [add: ’l-nas]).

17I did not find it in the relevant sections of the Mudawwana.


this incident and the lack of remarks by any later Muslim scholar makesit clear that the Christian story was unknown among medieval Muslims.

By contrast, the legend did not lose its popularity among Christiansafter the ninth century. Sundry versions of it are known from EuropeanLatin lives of Muh.ammad, especially from texts written from the twelfthcentury onwards, when European settlement in the Outremer acceleratedthe flow of Oriental Christian legends about Muh.ammad and the rise ofIslam to Europe. In them, Muh.ammad’s end becomes ever more ghastly;in many versions, his corpse is devoured by pigs instead of dogs, and insome, he dies torn apart by pigs.18

Although no anti-Islamic polemical writings produced by Christiansliving in the central Islamic lands after the ninth century included thestory (the two later Arabic translations of the Christian Bah. ıra leg-end tacitly omitted it) we can be fairly certain that it remained partof the Oriental Christian oral tradition. First, a distant but recogniz-able variant of the legend, which appears in the Apocalypse of Peter,a Christian Arabic text that probably originated in late ninth-centurySyria or Mesopotamia in the Syriac tradition, is also attested in theJudaeo-Arabic commentary that Yefet ben ↪Eli, a Karaite Jew who livedin Palestine in the late tenth century, wrote on the book of Isaiah.19

“Peter, verily I say to you that after the death of the Son of PerditionI will send the loathsome beast to him to dig him out from his graveand devour his flesh,” said Jesus to Peter according to the Apocalypse ofPeter.20 In Yefet’s commentary we read, “They removed him from hisgrave, and the lions ate him. Nothing remained from him save his heel.They took it and buried it, and said, ‘This is the grave of the man ofthe spirit’.”21 The most probable explanation of this similarity is thatthe legend was part of both Christian and Jewish lore about the rise

18Kohlberg, “Western accounts,” pp. 165–166; Tolan, “Un cadavre mutile,” pp.55–58 = idem, “A mangled corpse,” pp. 21–23; Tolan, Saracens, 142–143; Daniel,Islam and the West, pp. 125–129.

19For the Apocalypse of Peter, see Mingana (ed.), “Apocalypse of Peter,” p. 323(Karshunı text), and p. 254 (English translation); for the reference to the legend inYefet’s commentary, see Vajda, “Un vestige oriental,” pp. 177–179. On the datingand provenance of the Apocalypse of Peter, see Roggema, “Biblical exegesis,” pp.137–138.

20Mingana (ed.), “Apocalypse of Peter,” p. 323 (h. aqqan aqulu laka ya fat.rusa in-nahu idha tuwuffiya bnu ’l-halaki ursiluhu ’l-h. ayawana ’l-samja h. atta yanbushahumin qabrihi wa-ya↩kula lah. mahu). I have slightly modified Mingana’s English trans-lation (ibid., p. 254).

21Vajda, “Un vestige oriental,” p. 178. Yefet’s commentary on Isaiah remainsunedited; the passage in question (part of his exegesis of Isaiah 14:19) is transcribedibid., p. 178, n. 5. For the “man of the spirit,” see Hosea 9:7 (ibid., p. 178, n. 4).Although Yefet wrote his commentary in Judaeo-Arabic, most of this passage is inHebrew, probably out of cautiousness.


of Islam in the medieval Islamic milieu. Second, from the twelfth cen-tury onward, some authors of European Latin lives of Muh.ammad whoincluded the legend in their works refer to Oriental Christians as theirinformants on Muh.ammad’s life.22 Third, modifications of the narrativein late manuscripts of the ninth century Christian works exhibit variantsthat attest to the copyists’ familiarity with the story independently oftheir Vorlage.23

Whichever version one is acquainted with, the failed-resurrection leg-end at first sight seems but a malicious slander invented by Christians.Most writings that contain it are polemical, their tone often acrid andscathing. It reads as the inversion of the Christian story of Christ’sresurrection. The legend appears to have been assembled from literarytopoi, suitable for the polemical purposes of the Christian authors. Itcomes as no surprise that all scholars discussing the story have dismissedit in its entirety as a malignant Christian fantasy.24

However preposterous the Christian legend of Muh.ammad’s deathappears, the Christians did not invent it: they borrowed almost all of itsmotifs from the early Islamic tradition. In the rest of this paper, I willdiscuss each motif separately, and show that three out of four25 of themwere known among Muslims in the middle of the eighth century, theprobable terminus ad quem of the formation of the Christian narrative.

22For example, Adelphus (see Tolan, Saracens, pp. 138, 142). Note also Gautierde Compiegne’s ultimate source, the “Saracen convert to Christianity” (ibid.). Theseversions of the legend are, as mentioned above, often at variance with those knownfrom the ninth-century Middle East, and it cannot be ascertained which motifs orig-inated among Oriental Christians, and which among Europeans. Beginning with theCrusades, the European Latin versions exercised their own influence in the MiddleEast, which further complicates the question of sources. For example, see the twelfth-or thirteenth century Armenian abridgement of a Latin text (“[Extrait] de l’Histoiredes Latins”) in Macler, “Un document armenien,” pp. 287–295.

23One of the five manuscripts used for the edition of the West-Syrian recensionof the Christian Bah. ıra legend (Mingana Syr. 71; undated) exhibits several signifi-cant variant readings. For instance, the copyist inserts that guards were placed byMuh. ammad’s corpse (see Roggema, The legend of Sergius Bah. ıra, p. 334, n. 32); adetail unattested in any other version of the Christian Bah. ıra legend, but familiarfrom the Istoria and the Adnotatio. Mxit‘ar of Ani (see Appendix, pp. 159–162)might also have known the legend independently from his source. The Qashun doc-ument blames “the drowsy disciples” for the dogs’ defilement of the corpse; Mxit‘archarges “the guards” instead. Guards are never mentioned in the Qashun document(see Thomson, “Muh. ammad,” p. 850).

24See, for example, Hoyland, Seeing Islam, p. 514; Kohlberg, “Western accounts,”pp. 166–167; Wolf, “The earliest Latin lives,” p. 90; Daniel, Islam and the West, pp.19, 125–129; Tolan, Saracens, pp. 92–93; and idem, “Un cadavre mutile,” pp. 56–57= idem, “ A mangled corpse,” pp. 21–22.

25The four motifs are the following: (1) Muh. ammad’s resurrection and ascensionto heaven after his death, (2) the three-day delay in his burial, (3) the putrefactionof his corpse, (4) the dogs’ mauling of his corpse.


I will begin with discussing h. adıths about Muh.ammad’s resurrection andascension to heaven after his death, continue with the debate among earlyMuslims about the time elapsed between Muhammad’s death and burialas well as the state of his corpse by the time of his burial. Of the fourmotifs of the Christian legend, only the dogs’ mauling of Muh.ammad’scorpse appears to be unattested in the early Muslim tradition.

Muh. ammad’s ascension after his death in the Islamictradition

Regarding the first motif of the story, Muh.ammad’s foretelling of his ownresurrection, it is beyond doubt that the Christian polemicists borrowedit from the Muslim tradition. A comparison of two texts will show this.The first is Christian: “Furthermore, even more hideous and revoltingthan this was that he (Muh.ammad) used to say to them in his life andcommend to them that when he died they should not bury him becausehe would ascend to heaven as Christ, Lord of the World, ascended, andthat he is so precious to God that He would not leave him on the earth formore than three days (wa-annahu akramu ↪ala Allahi min an yatrukahu↪ala ’l-ard. i akthara min thalathati ayyamin),” begins the story in theApology of al-Kindı.26 The second text is Muslim: “I am so preciousto God that He will not leave me in the earth after three (days) (anaakramu ↪ala Allahi min an yatrukanı fı ’l-ard. i ba↪da thalathin),” saidMuh.ammad according to a rare h. adıth.27 Although the Christian andthe Muslim traditions differ in the time of the anticipated, the formerputting it before the burial (“on the earth”), and the latter after it (“inthe earth”), such a close correspondence in both content and wordingcan hardly be accidental. The author of the Apology clearly refers to aversion of this h. adıth.

It would be interesting to trace the transmission history of the h. adıthin order to learn when and where it circulated, and thus to gain a better

26See Tartar, Dialogue islamo-chretien, p. 92.27See Karajikı, Kanz al-fawa↩id, vol. 2, p. 140; Juwaynı, Nihayat al-mat.lab, vol.

3, p. 66; Ghazalı, al-Durra al-fakhira, p. 42. The tradition is attested in severalvariants, without significantly changing the meaning: sometimes ↪inda (“in the eyesof”) is read instead of ↪ala, rabbı (“my Lord”) instead of Allah, yada↪anı (“leavesme”) instead of yatrukanı, and qabrı (“my tomb”) instead of al-ard. . Al-Ghazalı’sinterpretation of ba↪da thalath as “after three decades” is clearly secondary. As otherrelated traditions, quoted below, explicitly refer to days — two days, forty days, andhalf a day — there is no reason to look for other meanings. Moreover, ba↪da thalathis attested in the sense of “after three days” in another context (see below).


understanding of its relationship with the Apology, but I was unable torecover it. The h. adıth is attested only in a few relatively late writings;none of them quotes it with isnad. The oldest surviving works whichmention it were written by the Imamı Shı↪ı al-Karajikı (d. 449/1057), andtwo Sunnıs, ↪Abd al-Malik al-Juwaynı (d. 478/1085) and al-Ghazalı (d.505/1111).28 Al-Juwaynı ascribes it to Abu ↪Alı al-Sinjı, a scholar of theprevious generation,29 and says that it was related also with the phrase“more than two days (akthara min yawmayni).” Al-Rafi↪ı (d. 623/1226)also cites the h. adıth in his al-Sharh. al-kabır.30 The authors of latercompilations about the traditions in al-Sharh. al-kabır, al-Zarkashı (d.794/1392), Ibn al-Mulaqqin (d. 804/1401) and Ibn H. ajar (d. 852/1449),unsuccessfully tried to trace its isnad.31 They could only refer to theh. adıth in older works; the earliest scholar named in this connectionis al-Azraqı, possibly Ah.mad b. Muh.ammad b. al-Walıd al-Azraqı (d.222/837) who is best known as the main source of his grandson’s AkhbarMakka.32 All we can conclude from the information they give is thatthe tradition once circulated in at least two versions, and was possiblyknown in the early ninth century.33

There are further h. adıths (I will call them “ascension traditions”)that corroborate an early Muslim belief in Muh.ammad’s ascension to

28See previous note.29He died in the 430s (1038 or later); see, for example, Ibn Khallikan, Wafayat

al-a↪yan, vol. 2, p. 115.30See Rafi↪ı, al-Sharh. al-kabır, vol. 2, p. 445.31See Ibn al-Mulaqqin, al-Badr al-munır, vol. 5, pp. 283–292; Ibn H. ajar, Talkhıs

al-h. abır, vol. 2, p. 293–294. The work of al-Zarkashı is lost or has not been printedyet, but his remarks on the h. adıth are quoted in Suyut.ı, al-La↩alı al-mas.nu↪a , vol.1, p. 285. Ibn al-Mulaqqin refers, without name, to a fourteenth century author ofa work on the prophets’ lives in their graves as mentioning the h. adıth, also withoutan isnad (wa-dhakarahu ba↪d. u man adraknahu. . . fa-lam ya↪zuhu); see al-Badr al-munır, vol. 5, p. 283. Even later authors, such as al-Suyut.ı or al-Zurqanı, refer tothe h. adıth, but they merely reiterate what the earlier ones said.

32Mentioned by al-Zarkashı who is quoted in Suyut.ı, al-La↩alı al-mas.nu↪a, vol. 1,p. 285. On Azraqı, see art. “al-Azrak. ı Abu ’l-Walıd Muh. ammad b. ↪Abd Allah b.Ah. mad,” s.v. Ibn H. ajar refers to Ibn Abı Layla (Muh. ammad b. ↪Abd al-Rah. man IbnAbı Layla al-Ans.arı, Abu ↪Abd al-Rah. man al-Kufı, d. 148/765–6, see Mizzı, Tahdhıbal-kamal, vol. 25, pp. 622–628) as a possible source of the h. adıth, but it seems thatIbn H. ajar’s only reason for doing so is another h. adıth with somewhat similar contentthat he quoted just before this one and that was transmitted by Ibn Abı Layla (seeIbn H. ajar, Fath. al-barı, vol. 7, p. 296).

33Two more traditions, expressed in words very similar to this, but related toeschatology also occur: “I am so precious to God that He will not leave me in theearth for a thousand years (ana akramu ↪ala Allahi min an yatrukanı fı ’l-turabi alfa↪amin),” and “I am so precious to God that He will not leave me under the earth fortwo hundred years (ana akramu ↪ala Allahi min an yada↪anı tah. ta ’l-ard. i mi ↩atay↪amin)” (see S. aghanı, Mawd. u↪at, p. 44; ↪Ajlunı, Kashf al-khafa↩, vol. 1, p. 161 and p.231; Ibn Kathır, al-Bidaya wa-’l-nihaya, vol. 5, p. 66). Both are rejected as forgeries.


heaven after his death and show that it existed already in the mid-eighth century. They clearly express the same belief as the ana akramutradition, but differ in significant details. All of them refer to the ascen-sion of prophets in general to heaven, not specifically to Muh.ammad’s.It is, however, likely that early Muslims created and circulated theseh. adıths out of interest in Muh.ammad’s, and not an earlier prophet’s,postmortem fate. On the one hand, the contexts in which some of theascension traditions are quoted directly connect them to Muh.ammad’sdestiny; on the other hand, many other h. adıths that make general state-ments about prophets clearly intend to say something primarily aboutMuh.ammad.34 Also, their wording shows less resemblance to the Chris-tian texts. Not all of them speak about ascension to heaven three daysafter death: according to some of them, the ascension took place fortydays (one variant does not specify the time the prophets remain in thegrave). Instead of having Muh. ammad speak in first person, they areascribed to later generations of Muslims. But the ascension traditionsare somewhat better documented than the ana akramu tradition, inso-far as they are at least quoted with isnads. Although all except oneare attested with a single isnad which limits the extent to which theirtransmission history can be reconstructed, some of it can be traced withcertainty.

“It is well-known,” says Ibn al-Mulaqqin in his discussion of the anaakramu tradition in al-Badr al-munır, “that the wall of the Prophet’stomb collapsed during the caliphate of al-Walıd b. ↪Abd al-Malik b.Marwan and the governorship of ↪Umar b. ↪Abd al-↪Azız over Medina,[and] a foot appeared to them. They dreaded that it might be the footof the Messenger of God. Its matter appalled them, and they were over-come with fear, until Sa↪ıd b. al-Musayyab related to them, ‘The corpsesof the prophets, may God bless them, do not remain in the earth morethan forty days, then they ascend.’ Salim b. ↪Abd Allah b. ↪Umar b.al-Khat.t.ab also came, and recognized in it the foot of his grandfather,↪Umar.”35 As far as I know, this is the only version of the story aboutMuh.ammad’s collapsed tomb that interweaves an ascension tradition,

34Compare, for example, the following h. adıths, “Each and every prophet tendedsheep (ma min nabiyyin illa qad ra↪a ’l-ghanama),” “No prophet dies until he is giventhe choice (between this world and the hereafter) (ma min nabiyyin yamutu h. attayukhayyaru),” and “No prophet is buried except where he dies (ma tawaffa Allahunabiyyan qat.t.u illa dufina h. aythu tuqbad. u ruh. uhu)” in Ibn Sa↪d, T. abaqat, vol. 1/1,pp. 79–80; vol. 2/2, pp. 28, and 71.

35Ibn al-Mulaqqin (d. 804/1401), al-Badr al-munır, vol. 5, p. 285 ([i ]nna juthatha’l-anbiya↩i s.alawatu Allahi ↪alayhim la tuqımu akthara min arba ↪ına yawman fı ’l-ard. ithumma turfa↪u).


and it is clearly a composite of the two different stories.36 It is not theonly version of ascension traditions connected to Muh.ammad’s presencein his grave, however. Two further variants are attributed to Sa↪ıd b.al-Musayyab who utters them, disapprovingly, when he observes peo-ple visiting Muh.ammad’s tomb. “Prophets possessing fortitude do notstay [in the earth] beyond forty days before they ascend [to heaven]; theProphet of God did not stay in the earth longer than forty days beforehe ascended,” said Sa↪ıd according to one of these.37 “No prophet re-mains in the earth for more than forty days,” he said according to theother.38 A very similar variant of the last one appears, without con-nection to Muh.ammad’s grave or the pilgrimage to it, in al-Bayhaqı’s(d. 458/1066) Kitab ma warada fı h. ayat al-anbiya↩ ba↪d wafatihim.39

All four versions are associated with Sa↪ıd b. al-Musayyab, the famousMedinan scholar of the late seventh and early eighth century.

If we trusted the attribution of these sayings to Sa↪ıd, possibly theironly common link,40 we could conclude with certainty that in the lateseventh or early eighth century some Muslims, at least in Medina, be-lieved that Muh.ammad had risen from the dead and ascended to heaven.We should not, however, rush to ascribe this opinion to Sa↪ıd, sinceelsewhere he is said to have been holding that Muh.ammad lives not

36According to other versions about the collapsed tomb, ↪Umar’s horror abatedafter ↪Urwa b. al-Zubayr (or Salim b. ↪Abd Allah b. ↪Umar b. al-Khat.t.ab, or Salim’sbrother, ↪Abd Allah) recognized the foot as ↪Umar b. al-Khat.t.ab’s. See Ibn Sa↪d,T. abaqat, vol. 3/1, p. 268; Ibn Kathır, al-Bidaya wa-’l-nihaya, vol. 9, p. 75; Samhudı,Khulas.at al-wafa↩, vol. 1, p. 322; and elsewhere.

37I found the only reference to this one in Samhudı (d. 911/1506), Khulas.at al-wafa↩, vol. 1, pp. 115–116 (innahu la yabqa nabiyyun min ulı ’l-↪azmi fawqa arba ↪ınalaylatan h. atta yurfa↪u inna nabiyya Allahi lam yabqa fı ’l-ard. i fawqa arba ↪ına laylatanh. atta rufi ↪a). The expression “possessing fortitude” is borrowed from Qur↩an 46:35.This tradition thus restricts the number of prophets who ascended to heaven fortydays after their death. Opinions vary how many and which prophets were “possessingfortitude,” ranging from all of them to only a handful, but the lists always end withMuh. ammad. See, for example, T. abarı, Tafsır, vol. 26, p. 37; Tha↪labı, Tafsır, vol. 9,pp. 24–26. I am thankful for the Qur↩anic reference to Professor Michael Cook.

38The earliest quotation of this one is ↪Abd al-Razzaq (d. 211/827), Mus.annaf, vol.3, p. 383 (ma makatha nabiyyun fı ’l-ard. i akthara min arba ↪ına yawman).

39Bayhaqı, H. ayat al-anbiya↩ (1), pp. 29–30; idem, H. ayat al-anbiya↩ (2), pp. 76–77(“No prophet remains in his grave for more than forty days before he ascends,” mamakatha nabiyyun fı qabrihi akthara min arba ↪ına laylatan h. atta yurfa↪u). The firstedition’s fı qabr instead of fı qabrihi must be a typographical error.

40The isnad of the first is Sufyan al-Thawrı, an unnamed shaykh, and Sa↪ıd b.al-Musayyab; of the second, ↪Abd al-Razzaq, Sufyan al-Thawrı, Abu ’l-Miqdam, andSa↪ıd b. al-Musayyab. Al-Samhudı does not give the full isnad, but attributes hisaccount to al-Minhal b. ↪Amr (“From al-Minhal b. ↪Amr [who said], ‘I was besideUmm Salama’s room with Sa↪ıd b. al-Musayyab. . . Sa↪ıd said. . . ”’). Ibn al-Mulaqqindoes not name any transmitters.


in heaven, but in his grave.41 Even if the attribution of the ascensionh. adıths to Sa↪ıd is apocryphal, their circulation can still be dated as earlyas the mid-eighth century because al-Bayhaqı says that his version wasincluded in the Jami ↪ of Sufyan al-Thawrı (d. 161/778).42 Whichever ofSufyan’s Jami ↪s al-Bayhaqı meant (he wrote a large and a small one), itis lost today, but there is no reason to doubt al-Bayhaqı’s statement. AsSufyan al-Thawrı was a Kufan scholar, the h. adıth must have circulatedin Kufa. Its association with the town is strengthened by the two tra-ditionists mentioned in the isnad of other versions as transmitting fromSa↪ıd b. al-Musayyab, Abu ’l-Miqdam and al-Minhal b. ↪Amr. Both ofthem were Kufans.43

As we saw, the earliest transmitter mentioned in connection withthese four ascension traditions is Sa↪ıd b. al-Musayyab, a Successor al-though their attribution to him is dubious. I found only one Sunnı h. adıthexpressing the same idea with a fuller isnad. “The prophets are not leftin their graves after forty nights, but are praying before God, may Hebe exalted and glorified, until the horn is blown,” said Muh.ammad ac-cording to a tradition quoted in al-Bayhaqı’s tract.44Its first transmitteris said to have been Anas b. Malik, Muh.ammad’s servant, who latersettled in Bas.ra, the second Thabit al-Bunanı (Bas.ran, d. 120s/740s),and the third Ibn Abı Layla (Kufan, d. 148/765–6).45 Since this is theonly known isnad of the tradition,46 its origins cannot be identified withany certainty.47 Still, Ibn Abı Layla belonged to the same generation

41It is related that Sa↪ıd b. al-Musayyab heard the adhan, or the noise of somebodypraying, or mumbling (hamhama) inside Muh. ammad’s tomb during the battle ofH. arra (63/683); see Darimı, Sunan, vol. 1, pp 227–228; Abu Nu↪aym, Dala↩il al-nubuwwa, p. 496; Ibn Sa↪d, T. abaqat, vol. 5, pp. 97–98.

42On him, see “Sufyan al-Thawrı,” EI2, s.v.43On Abu ’l-Miqdam al-H. addad, Thabit b. Hurmuz, Kufan as Sufyan (dates un-

known), and on al-Minhal b. ↪Amr al-Asadı, also a Kufan (dates are also unknown),see Mizzı, Tahdhıb al-kamal, vol. 4, pp. 380–381, and vol. 28, pp. 568–572.

44See Bayhaqı, H. ayat al-anbiya↩ (1), p. 29 (al-anbiya↩u la yutrakuna fı quburihimba↪da arba ↪ına laylatan wa-lakinnahum yus.alluna bayna yadayi Allahi ↪azza wa-jallah. atta yunfakhu fı ’l-s. uri).

45For Ibn Abı Layla see above, note 32; for Thabit b. Aslam al-Bunanı, AbuMuh. ammad al-Bas.rı, see Mizzı, Tahdhıb al-kamal, vol. 4, pp. 342–349. The iden-tity of the fourth transmitter, Isma↪ıl b. T. alh. a b. Yazıd, might also be relevant, butI was unable to identify him.

46The only author who may have given the h. adıth independently from al-Bayhaqıis al-Daylamı (see Firdaws, vol. 1, p. 222), but he ascribes it only to Anas b. Malik,omitting the full isnad. Later authors always quote the h. adıth from al-Bayhaqı.

47Another similar tradition, “No prophet dies and remains in his tomb except forforty days (ma min nabiyyin yamutu fa-yuqımu fı qabrihi illa arba ↪ına s.abah. an),”is attested in a tenth-century and in an eleventh-century work, but the latter adds,“until his spirit is returned to him (h. atta yuraddu ilayhi ruh. uhu),” fundamentallychanging the meaning. Since the early transmitters in the isnad are identical in both


of Kufan scholars as Sufyan al-Thawrı which supports the link of theascension traditions to eighth century Kufa.

Since only four additional ascension traditions are attested, two Sunnıand two Imamı Shı↪ı ones, it is worth quoting all of them here. One ofthem is connected to Kufa, similarly to the previously cited ones, whilethe transmission history of the rest cannot be reconstructed. One ofthe two Sunnı ascension traditions is a variant of Anas b. Malik’s h. adıth,which extends the privilege of early ascension to heaven to more groups ofpeople and leaves the number of days to be spent in the grave unspecified.It is quoted in Daylamı’s Firdaws: “Ten [groups of people] are not leftin their graves, but are praying before God, may He be exalted andglorified, until the horn is blown: the prophets, the martyrs, those whocall to prayer, those who obey [the call to prayer], the one who dies onthe way to Mecca, the woman who dies in childbirth, those who repenttheir sins, the one who serves the Muslims in obedience to God, may Hebe exalted and glorified, those who pray at night while the people areasleep, and those who have mercy over the poor of my community.”48 Itstransmission history cannot be reconstructed.49 The other Sunnı saying,“God does not leave a prophet in his grave for more than half a day”was cited by the H. anbalı Abu ’l-H. asan Ibn al-Zaghunı (d. 527/1132),50

but I was unable to trace further information about it. Al-Shawkanı (d.1255/1839) might have been familiar with other variants of the ascensiontraditions too; he refers to them in one of his works, and his wording isdifferent from other authors.51 Perhaps further Sunnı traditions of this

cases, it is impossible to decide which version is the original. For this h. adıth, seeAbu Nu↪aym, H. ilyat al-awliya↩ (with typographical errors), vol. 8, p. 333; and IbnH. ibban, Kitab al-majruh. ın, vol. 1, p. 285.

48Daylamı, Firdaws, vol. 3, p. 35 (↪ashara la yutrakuna fı quburihim wa-lakinnahum yus.alluna bayna yadayi Allahi ↪azza wa-jalla h. atta yunfakhu fı ’l-s. uri’l-anbiya↩u wa-’l-shuhada↩u wa-’l-mu↩adhdhinuna wa-’l-mulabbuna wa-’l-mutawaffafı t.arıqi makkata wa-’l-mar ↩atu tamutu fı nifasiha wa-’l-ta↩ibuna mina ’l-dhunubiwa-khadimu ’l-muslimına fı t.a↪ati Allahi ↪azza wa-jalla wa-’l-mus.alluna bi-’l-layliwa-’l-nasu niyamun wa-’l-mutarah. h. imuna ↪ala fuqara↩i ummatı). The only editionof the text has the erroneous “h. atta yunfakhu fı ’l-s.adri” instead of “fı ’l-s. uri.”

49The h. adıth is ascribed to ↪Abd Allah b. Ja↪far (b. Abı T. alib); according to thefootnote, another manuscript attributes it to ↪Abd al-Rah. man b. Ja↪far (ibid.). Noneof this helps to trace the origins of the tradition.

50So it is said in Ibn ↪Abd al-Hadı, al-S. arim al-munkı, p. 273, and in Suyut.ı, al-La↩alı al-mas.nu↪a, vol 1, p. 285 (inna Allaha la yatruku nabiyyan fı qabrihi aktharamin nis.fi yawmin). Ibn al-Zaghunı’s Id. ah. (apparently his only work published so far)does not contain the tradition.

51Al-Shawkanı, Nayl al-awt.ar, vol. 5, p. 178 (“. . . it has come down that theprophets are not left in their graves beyond three [days], and it was also related[that] beyond forty [days]” (qad warada anna ’l-anbiya↩a la yutrakuna fı quburihimfawqa thalathin wa-ruwiya fawqa arba ↪ına). Only al-Samhudı quotes an ascensiontradition with fawqa arba ↪ına instead of the usual ba↪da arba ↪ına (see above); the


kind also circulated once, but were subsequently forgotten.An Imamı Shı↪ı h. adıth, “No prophet or legatee remains in the earth

for more than three days before he ascends to heaven in his spirit, hisbones and his flesh. . . ,” ascribed to the sixth imam, Ja↪far al-S. adiq (d.148/765), appears in four late ninth- and tenth-century writings, withidentical isnads. The two earliest transmitters, Ziyad b. Abı ’l-H. alaland ↪Alı b. al-H. akam, were both Kufans, providing further support forthe circulation of the h. adıth in Kufa in the late eighth and early ninthcentury.52 Another Imamı Shı↪ı h. adıth, also attributed to Ja↪far al-S. adiq,claims that “the corpse of a prophet or of a prophet’s legatee does notremain in the earth for more than forty days.”53 The information Ifound about its early transmitters is insufficient and cannot support anyconclusions about its circulation.54

Although only a small corpus, the ascension traditions indicate abelief in Muh.ammad’s ascension to heaven after his death in some circlesin pre-classical Islam. The isnads point to Kufa as a place where thisbelief might have been more widely accepted than elsewhere, but theavailable information is too scarce to associate it exclusively with thistown. The bulk of all recorded Islamic traditions are of Kufan, Bas.ranand Medinan provenance. Therefore, while the isnad pattern of theascension traditions might be understood as evidence that the beliefin Muh.ammad’s ascension to heaven was held by more Muslims in Kufathan in Bas.ra or in Medina, it does not say anything about its diffusion inother regions of the Islamic world. The dating of the ascension traditionsis similarly problematic. On the basis of the isnads, we can be fairlycertain that such a belief was held already in the middle of the eighthcentury, but it is hardly possible to trace when it first appeared, or how

phrase fawqa thalathin does not occur elsewhere.52Abu Ja↪far al-Qummı, Bas. a↩ir al-darajat, vol. 2, p. 349; Ibn Quluya, Kamil al-

ziyarat, pp. 329–330; Ibn Babawayh, Faqıh, vol. 2, pp. 577–578; and Kulaynı, Kafı,vol. 4, p. 567 (ma min nabiyyin wa-la was.iyyin yabqa fı ’l-ard. i akthara min thalathatiayyamin h. atta yurfa↪u bi-ruh. ihi wa-↪az.mihi wa-lah. mihi ila ’l-sama↩i . . . ). The versionin Kamil al-ziyarat has some textual variants, without changing the meaning. Thefirst four transmitters in all four works are Ja↪far al-S. adiq, Ziyad b. Abı ’l-H. alal, ↪Alıb. al-H. akam b. al-Zubayr, and Ah. mad b. Muh. ammad b. ↪Isa al-Ash‘arı. Ziyad b. Abı’l-H. alal is named as transmitting from Ja↪far, ↪Alı b. al-H. akam from ↪Alı al-Rid. a (d.203/818) and his son, Muh. ammad Jawad Taqı (d. 220/835). On Ziyad and Ah. mad,see Najashı, Rijal, vol. 1, p. 390 and pp. 216–218; on ↪Alı, see T. usı, Rijal, p. 361.

53See T. usı, Tahdhıb al-ah. kam, vol. 6, p. 118 (la tamkuthu juththatu nabiyyin wa-lawas.iyyi nabiyyin fı ’l-ard. i akthara min arba ↪ına yawman).

54The first transmitter, ↪At.iyya al-Abzarı is listed among the disciples of Ja↪far al-S. adiq (see T. usı, Rijal, p. 260). I was unable to trace any information about ↪Amr b.Ziyad, the second transmitter. The third is referred to as al-↪Ala↩ b. Yah. ya, brother ofMughallis; he is Kufan, if he is identical with al-↪Ala b. Yah. ya al-Makfuf in Najashı,Rijal, vol. 2, p. 154.


long it continued to be accepted, and just how popular it was at anytime. With so many eighth-century works lost today, it could have beenmore widespread than it now seems to us.55

The ideas expressed in the ascension traditions resemble the Chris-tian stories about Muh.ammad’s expected resurrection and ascension toheaven. The ana akramu tradition was directly quoted by the authorof the Apology of al-Kindı. There is, however, a crucial difference: ac-cording to the Islamic traditions the resurrection and the ascension weresupposed to happen and did indeed happen after burial, while accordingto the Christian stories they were meant to take place without burial,and eventually failed to do so. Were it for only the ascension tradi-tions it could be argued that the Christian legend of Muh.ammad’s deathoriginated as the rejection of the eighth-century Muslim belief. Therecirculated, however, further h. adıths that the Christians drew on whenconstructing their account of Muh.ammad’s death.

Muh. ammad’s belated burial in the Islamic tradition

According to Ibn Hisham, Muh.ammad died in the late morning of aMonday in Rabı↪ al-Awwal, and was buried “in the middle of the nightof Wednesday,”56 that is, on Tuesday night, according to our reckon-ing of time. Islamic tradition, both Sunnı and Imamı Shı↪ı, agreesabout the day of Muh.ammad’s death, but is divided over the day of hisburial. Some Sunnı h. adıths claim that Muh.ammad was interred yawma’l-thulatha↩ (between the sunsets of Monday and Tuesday), while otherSunnı and apparently all Shı↪ı h. adıths maintain that it happened yawma’l-arbi ↪a↩ (between the sunsets of Tuesday and Wednesday).57 There

55The terror that seized ↪Umar b. ↪Abd al-↪Azız when he thought that he sawMuh. ammad’s foot behind the collapsed wall of his grave might also be related to thisbelief. Even though only one version connects Muh. ammad’s ascension to the story,the question remains as to why would ↪Umar be awed to see Muh. ammad’s foot, butcalm down when told it is ↪Umar’s, unless the former was not supposed to be in thegrave at all.

56Ibn Hisham, al-Sıra al-nabawiyya, vol. 1/2, pp. 1009–1011, 1020 (wasat.a ’l-laylilaylata ’l-arbi ↪a↩i or jawfa ’l-layli min laylati ’l-arbi ↪a↩i).

57See the remarks of al-T. abarı and Ibn Kathır to this effect (T. abarı, Ta↩rıkh, vol.1/4, pp. 1815, 1830; Ibn Kathır, al-Bidaya wa-’l-nihaya, vol. 5, pp. 193, 206); and thelists of traditions in Ibn Sa↪d, T. abaqat, vol. 2/2, pp. 57–59, 78–79; T. abarı, Ta↩rıkh,vol. 1/4, pp. 1831–1832, 1832–1833, 1837 (two h. adıths); and Ibn Kathır, al-Bidayawa-’l-nihaya, vol. 5, pp. 193–194, 205–206. The section of Majlisı, Bih. ar al-anwar onMuh. ammad’s death (vol. 22, pp. 503–549), on the other hand, only quotes traditionsthat refer to yawma l-arbi ↪a↩ as the day of burial, and mentions one tradition thatplaces the death on a Friday (ibid., p. 521).


circulated at least one tradition placing Muh.ammad’s burial yawma ’l-khamıs (between the sunsets of Wednesday and Thursday).58 Traditionsfurther vary with regard to the time of the day when Muh.ammad wasinterred.

The first two opinions are both attested in the middle of the eighthcentury; although the yawma ’l-arbi ↪a↩ tradition seems to have gainedwider currency by that time, the other one was also known.59 Muh. am-mad b. Ish. aq (d. ca. 150/767) is a firmly supported common link in theisnads of the yawma ’l-arbi ↪a↩ tradition; several traditionists transmittedit from him.60 An older common link, Makh. ul al-Shamı (d. 112–8/730–7), supported by two transmitters from him, makes it likely that theyawma ’l-arbi ↪a↩ tradition circulated in Syria already in the first decadesof the eighth century or earlier.61 I did not find a similarly old commonlink for the yawma ’l-thulatha↩ tradition, but it cannot be excluded thatit was disseminated just as early as its rival. I could not trace the originsof the yawma ’l-khamıs tradition.

A small group of Iraqi traditions claiming that Muh.ammad was in-terred only when the decomposition of his corpse became visible ap-pears to correspond to a later day of burial. According to a Kufanh. adıth, by the time Muh.ammad’s body was committed to earth, itscolor had changed.62 According to another Kufan tradition, Muh.ammad

58Al-Diyarbakrı, the only biographer of Muh. ammad who mentions the tradition,quotes it from the Tafsır al-Zahidı and the Kanz al-↪ibad, two thirteenth centuryworks (see Diyarbakrı, Ta↩rıkh al-khamıs, vol. 2, pp. 172), if indeed the author of thefirst is Mukhtar b. Mah. mud al-Ghazmını al-Zahidı (d. 658/1259–60). I did not findany tafsır attributed to him.

59See the remarks of Ibn Kathır, al-Bidaya wa-’l-nihaya, vol. 5, pp. 205–206. Inaddition to Ibn Ish. aq, he refers to Sulayman al-Taymı, Ja↪far al-S. adiq and Musa b.↪Uqba by name as deciding for yawma ’l-arbi ↪a↩, and mentions al-Awza↪ı and Sufyanal-Thawrı as holding the other opinion.

60Most versions mention Ibn Ish. aq as their transmitter. For these, see Ibn Hisham,al-Sıra al-nabawiyya, vol. 1/2, p. 1020; Baladhurı, Ansab al-ashraf (1), vol. 1, pp.657, 661–662; T. abarı, Ta↩rıkh, vol. 1/4, pp. 1832–1833, 1837; and elsewhere.

61Abu ↪Abd Allah Makh. ul al-Shamı, a Damascene transmitter who died three orfour decades earlier than Ibn Ish. aq (ca. 112–8/730–7). Ibn Kathır and al-Baladhurıgive two yawma ’l-arbi ↪a↩ traditions. Makh. ul is their oldest transmitter and theironly common link. For the first tradition, see above, n. 57; for the second, Baladhurı,Ansab al-ashraf (1), vol. 1, p. 657; on Makh. ul, see Mizzı, Tahdhıb al-kamal, vol. 28,pp. 464–475.

62 See Baladhurı, Ansab al-ashraf (2), vol. 1, p. 568 (dufina yawma ’l-thulatha↩ih. ına zaghat al-shamsu wa-taghayyara lawnuhu). This tradition is ascribed to theKufan Abu Mikhnaf (d. 154/774); on him, see “Abu Mikhnaf,” EI2, s.v. Not alleditions of Ansab al-ashraf contain the last phrase. In the one published in Damascusin 1996, the words are missing from the main text, and the editor explains in thefootnote that the phrase is effaced in the manuscript “because nobody transmittedthat” (see Baladhurı, Ansab al-ashraf [1], vol. 1, p. 657, n. 4).


was interred only when “death was apparent on him,” and his finger-nails turned greenish.63 Still another Kufan tradition claims that by thetime of Muh.ammad’s burial his corpse was bloated and his little fingerbent.64 A Bas.ran h. adıth similarly mentions the commencement of bloat-ing before burial.65 Finally, a Kufan h. adıth begins as follows, “when theProphet died, Abu Bakr was absent, and he came after three (days). Noone dared to uncover his face until his abdomen became ashen-colored.Abu Bakr uncovered his face. . . ”66

63 See Ibn Sa↪d, T. abaqat, vol. 2/2, p. 59 (lam yudfan rasulu Allahi h. atta ↪urifa’l-mawtu fıhi fı [sic] az.farihi ’khd. arrat). The oldest transmitter is a grandson of AbuBakr, Abu Muh. ammad al-Qasim b. Muh. ammad al-Qurashı al-Taymı (Medinan, d.ca. 101–12/719–31); he is followed by Abu ↪Abd Allah Jabir b. Yazıd b. al-H. arith al-Ju↪fı (Kufan, d. 128/745–6), Abu Muh. ammad Qays Ibn al-Rabı↪ al-Asadı (Kufan, d.ca. 165–8/781–5), and finally the Medinan al-Waqidı (d. 207/822) who later settledin Baghdad. On the first three transmitters, see Mizzı, Tahdhıb al-kamal, vol. 23, pp.427–436; vol. 4, pp. 465–472; and vol. 24, pp. 25–38; on the last, see “al-Waqidı,”EI2, s.v.

64See Ibn Sa↪d, T. abaqat, vol. 2/2, pp. 58–59 (turika rasulu Allahi ba↪da wafatihiyawman wa-laylatan h. atta raba qamıs.uhu wa-ru↩iya fı khins.irihi inthina↩). Thetransmitters are ↪Abd Allah al-Bahı, mawla of Mus.↪ab b. al-Zubayr (no dates orplaces are known for al-Bahı, but Mus.↪ab, the governor of Iraq, died in 72/691); Abu↪Abd Allah Isma↩ıl Ibn Abı Khalid al-Bajalı al-Ah. ması (Kufan, d. ca. 145–6/762–4);and Abu Sufyan Wakı↪ b. al-Jarrah. b. Malıh. al-Ru↩ası (Kufan, originally from Per-sia or Sogdia, d. ca. 196–7/811–3). For the three transmitters, see Mizzı, Tahdhıbal-kamal, vol. 16, pp. 341–342; vol. 3, pp. 69–76; and vol. 30, pp. 462–484; for Wakı↪,see also “Wakı↪ b. al-Djarrah. ,” EI2, s.v.; on Mus.↪ab, see “Mus.↪ab b. al-Zubayr,” EI2,s. v. The h. adıth is quoted elsewhere too.

65 See Ibn Sa↪d, T. abaqat, vol. 2/2, pp. 56–57 (lamma qubid. a rasulu Allahi i’tamaraas.h. abuhu fa-qalu tarabbas. u bi-nabiyyikum la↪allahu ↪urija bihi [qala] fa-tarabbas. u bihih. atta raba bat.nuhu. . . ). The transmitters are al-H. asan al-Bas.rı (d. 110/728); AbuSahl ↪Awf b. Abı Jamıla al-↪Abdı al-Hajarı (Bas.ran, of Iranian origin d. 146–7/763–5);Abu Nas.r ↪Abd al-Wahhab b. ↪At.a↩ al-Khaffaf al-↪Ijlı (Bas.ran, also lived in Baghdad;d. 200–6/815–22). For the first, see “al-H. asan al-Bas.rı,” EI2, s.v.; for the latter two,see Mizzı, Tahdhıb al-kamal, vol. 22, pp. 437–441; and vol. 18, pp. 509–516.

66 See T. abarı, Ta↩rıkh, vol. 1/4, p. 1817 (lamma qubid. a al-nabiyyu kana Abu Bakringha↩iban fa-ja↩a ba↪da thalathin wa-lam yajtari ↩ ah. adun an yakshifa ↪an wajhihi h. attairbadda bat.nuhu fa-kashafa ↪an wajhihi...). I am not sure why The history of al-T. abarı(vol. 9, p. 185) translates bat.n as “exterior,” and interprets ba↪da thalath as referringto hours. In view of the context and the h. adıths quoted above, it seems more likelythat the expression refers to days. See also Ibn Abı ’l-H. adıd who had no doubt thatal-T. abarı’s ba↪da thalath here means “after three days” (see Sharh. , vol. 13, pp. 35–37). The transmitters are Ibrahım al-Nakha↪ı (Kufan, d. ca. 96/714–5); a certain AbuAyyub; Abu Ma↪shar Ziyad b. Kulayb al-Tamımı al-H. anz.alı (Kufan, d. 110/728–9 or119/737); Abu Hisham (al-)Mughıra b. Miqsam al-D. abbı (Kufan, d. ca. 132–6/749–54); Abu ↪Abd Allah Jarır b. ↪Abd al-H. amıd al-D. abbı al-Razı (originally from theregion of Isfahan, grew up in Kufa, then settled around Rayy; d. 188/804); and Abu↪Abd Allah Muh. ammad Ibn H. umayd al-Tamımı al-Razı (d. 248/862–3). For the firsttransmitter, see “Ibrahım al-Nakha↪ı,” EI2, s. v.; for the rest, see Mizzı, Tahdhıbal-kamal, vol. 9, pp. 504–506; vol. 28, pp. 397–403; vol. 4, pp. 540–551; and vol. 25,pp. 97–108. I was unable to identify Abu Ayyub.


The oldest transmitter named in the isnads of the foregoing traditionsis ↪Abd Allah al-Bahı who probably died in the late seventh or the earlyeighth century.67 The earliest transmitters mentioned in three othersdied in the early eighth century: Ibrahım al-Nakha↪ı (d. ca. 96/714–5),al-Qasim b. Muh.ammad (a grandson of Abu Bakr; d. ca. 101–12/719–31), and al-H. asan al-Bas.rı (d. 110/728). The fifth tradition is ascribedsolely to Abu Mikhnaf (d. 154/774).68 That none of the isnads goes backto a supposed eyewitness is trust-inspiring. The earliest traditionistsmentioned, or at least most of them, probably indeed transmitted theseh. adıths, thus indicating a circulation of these narratives in Iraq by thebeginning of the eighth century or earlier.

The contents of these Islamic traditions resemble the Christian legendof Muh. ammad’s death. Both mention that Muh.ammad’s burial tookplace on the third day following his death,69 and that during this timehis followers were anxiously waiting for something to happen. Bothclaim that when Muh.ammad was finally interred, his body was in theprocess of decomposition. The Christian legend therefore did not invent,but borrowed from an Islamic narrative of Muh.ammad’s death, boththe outline of events (the delayed burial and the putrefied corpse) andpart of the explanation for it (from the ana akramu tradition). Theseh. adıths, however, each contain only one motif of the Christian narrative.Another Islamic tradition, a version of the story about ↪Umar’s denial ofMuh.ammad’s death, presupposes a conjunction of several elements thatalso occur in the Christian legend.

After Muh.ammad’s death, says Ibn Hisham, ↪Umar b. al-Khat.t.abaddressed the Muslims and denied Muh.ammad’s death. He accusedof hypocrisy those who claimed that Muh.ammad died, and threatenedthem with severe punishment after Muh. ammad would return. He com-pared the situation to the story of Exodus 32, and asserted that Mu-h. ammad “has gone to his Lord as Musa b. ↪Imran went and was hiddenfrom his people for forty days, returning to them after it was said that

67See above, n. 64.68For the traditionists, see above, nn. 62–63, 65–66.69Al-T. abarı, when summarizing the two opinions about the day of Muh. ammad’s

burial, contrasts the view that he was buried yawma ’l-thulatha↩ with the one that “hewas buried three days after his death (dufina ba↪da wafatihi bi-thalathati ayyamin),”clearly referring to the yawma ’l-arbi ↪a↩ tradition (Ta↩rıkh, vol. 1/4, p. 1830). Al-Fasawı also quotes a tradition according to which Muh. ammad “remained for threedays without burial (makatha thalathata ayyamin la yudfanu)” (al-Ma↪rifa wa-’l-ta↩rıkh, vol. 3, pp. 289–290). Some did not agree with calculating the period fromyawma ’l-ithnayn to yawma ’l-arbi ↪a↩ as three days. Ibn Kathır rejected this part ofthe tradition with indignation, adding that the right expression is that “he remained[unburied] for the rest of yawma ’l-ithnayn, the entire yawma ’l-thulatha↩, and wasburied on the night of yawma ’l-arbi ↪a↩.”


he had died. By God, the apostle will return as Moses returned. . . ”While ↪Umar was speaking, Abu Bakr arrived and immediately pro-ceeded to the house of ↪A↩isha to ascertain Muh.ammad’s death. Hethen tried to draw ↪Umar aside, but the latter would not listen to him.Abu Bakr nevertheless commenced his own speech. The Muslims, saysIbn Hisham, immediately came to listen to him, “People! If anyoneworshiped Muh. ammad, Muh.ammad is dead; if anyone worshiped God,God is alive, does not die.” He then recited a passage from the Qur↩an,“Muh.ammad is naught but a Messenger; Messengers have passed awaybefore him. Why, if he should die or is slain, will you turn about onyour heels? If any man should turn about on his heels, he will not harmGod in any way; and God will recompense the thankful.”70 Accordingto Ibn Hisham, the Muslims reacted as if they had never before heardthis passage.71

Several versions of the ↪Umar story appear in biographies of Muh. am-mad and h. adıth collections. The protagonists of most versions are ↪Umarand Abu Bakr. The speech of both men is heavily couched in Qur↩anic id-ioms, and Abu Bakr quotes Qur↩anic passages to prove that Muh.ammadhad to die like any other man. The story, as told by Ibn Hisham, cre-ates the impression that ↪Umar alone believed that Muh.ammad did notdie, and his attempt to convince other Muslims about this was nippedin the bud by Abu Bakr who arrived at an opportune moment. Mostversions paint a similar picture;72 at least one version, however, presentsthe events differently. In it, ↪Umar does not act alone, but with manyMuslims sharing his opinion.73

In this version of the story, it is ↪Abbas b. ↪Abd al-Mut.t.alib whoopposes ↪Umar. His speech strikingly differs from that of Abu Bakr.It alludes to a prolonged disagreement over the burial of Muh.ammad,the distressing physical symptom of his death, and the expectation ofMuh.ammad’s resurrection among the Muslims. ↪Abbas says, arguingagainst ↪Umar, “The Messenger of God has died. He is a mortal, and, as

70Qur↩an 3:144 (Arberry’s translation).71Ibn Hisham, al-Sıra al-nabawiyya, vol. 1/2, pp. 1012–1013 (dhahaba ila rab-

bihi kama dhahaba Musa bnu ↪Imrana fa-qad ghaba ↪an qawmihi arba ↪ına laylatanthumma rada↪a ilayhim ba↪da an qıla qad mata wa-wa-’llahi la-yarji ↪anna rasuluAllahi kama raja↪a Musa, and ayyuha ’l-nas innahu man kana ya↪budu Muh. ammadanfa-inna Muh. ammadan qad mata wa-man kana ya↪budu Allaha fa-inna Allaha h. ayyunla yamutu); translations from Guillaume, The life of Muh. ammad, pp. 682–683 (mod-ified).

72See, for example, Ibn Sa↪d, T. abaqat, vol. 2/2, pp. 53–56 (several versions);Baladhurı, Ansab al-ashraf (1), vol. 1, pp. 651–652, 654–656 (four versions); T. abarı,Ta↩rıkh, vol. 1/4, pp. 1815–1819 (three versions); ↪Abd al-Razzaq, Mus.annaf, vol. 5,pp. 301–302; Ibn Abı Shayba, Mus.annaf, vol. 7, pp. 427, 429.

73See Ibn Sa↪d, T. abaqat, vol. 2/2, p. 53.


it is with mortals, his odor changes. People, bury your Master. He is soprecious to God that He will not let him die twice. Would He let you dieonce, him twice? He is too precious to God for that. People, bury yourMaster. If it is indeed as you say nothing can prevent God from digginghim up from the earth. By God, the Messenger of God did not dieuntil he left the path plain and clear, allowed the lawful and prohibitedthe unlawful, married and divorced, warred and made peace. . . People,bury your Master.”74 On the basis of its isnads, the h. adıth probablycirculated in Bas.ra, in the middle of the eighth century at the latest.75↪Abbas’s speech in this h. adıth assumes an expectation of Muh.ammad’sresurrection on the part of ↪Umar and other Muslims instead of a denialof Muh.ammad’s death. If it assumed only Muh.ammad’s death it wouldnot argue that God would not allow his Prophet die twice.

A few unique traditions similarly hint that disbelief in Muh.ammad’s74Ibn Manz.ur, Mukhtas.ar ta↩rıkh Dimashq, vol. 2, pp. 385–386 (inna rasula Allahi

qad mata wa-innahu la-basharun wa-innahu ya↩sinu kama ya↩sinu ’l-basharu ayqawmu fa-’dfinu s. ah. ibakum fa-innahu akramu ↪ala Allahi min an yumıtahu imatatayna-yumıtu ah. adakum imatatan wa-yumıtuhu ithnatayn huwa akramu ↪ala Allahi mindhalika ay qawmu fa-’dfinu s. ah. ibakum fa-in yaku kama taquluna fa-laysa yu↪dhabu↪ala Allahi an yanjutha ↪anhu ’l-turaba inna rasula Allahi wa-llahi ma mata h. attataraka ’l-sabıla nahjan wad. ih. an fa-ah. alla ’l-h. alala wa-h. arrama ’l-h. arama wa-nakah. awa-t.allaqa wa-h. araba wa-salama. . . ay qawmu fa-’dfinu s. ah. ibakum). The sectionson Muh. ammad’s death are missing from from Shırı’s edition of Ibn ↪Asakir’s Ta↩rıkhmadınat Dimashq (apparently missing from all the extant manuscripts; see vol. 4,p. 394). For other variants of the h. adıth, see ↪Abd al-Razzaq, Mus.annaf, vol. 5, pp.300–301; Ibn Sa↪d, T. abaqat, vol. 2/2, pp. 53–54; Darimı, Sunan, vol. 1, pp. 220–222;and Baladhurı, Ansab al-ashraf (2), vol. 1, p. 567; ibid. (1), vol. 1, pp. 655–656 (thisedition replaces al-↪Abbas with Ibn ↪Abbas). Although much later than the others,I chose to translate Ibn Manz.ur’s text because it contains the same ideas as otherlonger versions of the h. adıth, and presents them in a more logical order.

75Ibn Manz.ur does not give a full isnad, only attributes the h. adıth to the Medinan↪Ikrima (d. ca. 105/723–4). ↪Abd al-Razzaq has Abu Bakr Ayyub b. Abı Tamımaal-Sakhtiyanı (Bas.ran, d. 131/748–9), and then Abu ↪Urwa Ma↪mar b. Rashid al-Azdıal-H. uddanı (Bas.ran, d. 150–4/767–71) transmit it from ↪Ikrima. The remaining threeh. adıth collections name Abu Isma↪ıl H. ammad b. Zayd al-Azdı al-Jahd. amı (Bas.ran,d. 179/795) as the third transmitter instead of Ma↪mar, and after H. ammad theyascribe the transmission to three different traditionists, respectively: Abu ’l-Nu↪manMuh. ammad (↪Arim) b. al-Fad. l al-Sadusı (Bas.ran, d. 223–4/837–9); Abu Ayyub Su-layman b. H. arb al-Azdı al-Washih. ı (Bas.ran, d. ca. 223–7/837–42); and a certainZayd b. Yah. ya al-Anmat.ı. Al-Baladhurı has Abu Zayd ↪Umar b. Shabba al-Numayrı(Bas.ran, d. 262/876) as the last transmitter, and inserts Ibn ↪Abbas (d. 68/687–8)as the first transmitter before ↪Ikrima, surely an instance of the backward growth ofisnads. For ↪Ikrima and Ibn ↪Abbas, see “↪Ikrima,” and “↪Abd Allah b. (al-)↪Abbas,”s.vv. For the other transmitters, see Mizzı, Tahdhıb al-kamal, vol. 3, pp. 457–464(Ayyub); vol. 27, pp. 303–312 (Ma↪mar); vol. 7, pp. 239–252 (H. ammad); vol. 26, pp.287–292 (Muh. ammad); vol. 11, pp. 384–393 (Sulayman); and vol. 21, pp. 386–390(↪Umar). I did not find any biography of Zayd b. Yah. ya al-Anmat.ı, but he is listedin Mizzı’s biography of ↪Umar b. Shabba as one of the traditionists from whom thelatter transmitted (see ibid., vol. 21, p. 387).


death was rampant in the Muslim community. A Medinan tradition de-picts the Muslim community as divided into two parties over the questionwhether Muh.ammad had died or not.76 According to a Bas.ran h. adıth,when Muh.ammad died, his Companions (as.h. abuhu) assembled and de-cided to wait with the burial because “perhaps he ascended” (la↪allahu↪urija bihi).77 According to another tradition, ↪Uthman also assertedthat Muh.ammad did not die. Unlike ↪Umar, however, ↪Uthman claimedthat Muh.ammad ascended to heaven just as Jesus did (rufi ↪a kama rufi ↪a↪Isa bnu Maryama).78 According to still another (possibly Medinan)h. adıth, “the people” (al-nas) denied that Muh.ammad died, and they be-lieved that he temporarily ascended to heaven similarly to Jesus (rufi ↪akama rufi ↪a ↪Isa bnu Maryama). These people, claims the tradition,threatened those who claimed that Muh.ammad died, and demandedthat Muh.ammad not be buried.79

Conclusion: le cadavre exquis?

We can thus conclude that most motifs of the Christian legend of Mu-h. ammad’s death were present in the Islamic tradition in the early eighthcentury, and it was not the Christians who invented them. The creativity

76Ibn Sa↪d, T. abaqat, vol. 2/2, p. 57 (akhbarana Muh. ammadu bnu ↪Umarah. addathanı al-Qasimu bnu Ish. aqa ↪an ummihi ↪an abıha al-Qasimi bni Muh. ammadibni Abı Bakrin aw ↪an Ummi Mu↪awiyata annahu lamma shukka fı mawti al-nabiyyiqala ba↪d. uhum mata wa-qala ba↪d. uhum lam yamut. . . ). The only Umm Mu↪awiya Ifound is Hind bint ↪Utba b. Rabı↪a who would be a rather unlikely transmitter in thisisnad. The other first transmitter is al-Qasim b. Muh. ammad b. Abı Bakr al-S. iddıqal-Qurashı al-Taymı, Abu Muh. ammad (Medinan, d. ca. 101–17/719–36), followed byhis daughter, Umm H. akım; then by her son, al-Qasim b. Ish. aq b. ↪Abd Allah b.Ja↪far b. Abı T. alib; finally by al-Waqidı. About Hind, see “Hind bt. ↪Utba,” EI2,s.v.; about al-Qasim b. Muh. ammad and al-Waqidı, see above, nn. 63–64. I foundUmm H. akım bint al-Qasim b. Muh. ammad b. Abı Bakr and his son mentioned in IbnQutayba, Ma↪arif, p. 208. He does not give any date or place. T. usı, Rijal, p. 271lists al-Qasim b. Ish. aq as a Medinan disciple of Ja↪far al-S. adiq (d. 148/765).

77Ibn Sa↪d, T. abaqat, vol. 2/2, p. 57; for the full text and the transmitters, seeabove, note 65.

78Baladhurı, Ansab al-ashraf (1), vol. 1, p. 655. The h. adıth is ascribed to al-Waqidı(wa-rawa al-Waqidiyyu fı isnadin lahu). I did not find it with a more detailed isnad.

79Ibn Sa↪d, T. abaqat, vol. 2/2, p. 57. The first transmitter is Abu Salama b. ↪Abdal-Rah. man b. ↪Awf al-Qurashı al-Zuhrı (Medinan, d. ca. 94–104/712–23); followed byZayd b. Abı ↪Attab (n.d.); then the unknown Maslama b. ↪Abd Allah b. ↪Urwa b. al-Zubayr; and al-Waqidı. For the first two transmitters, see Mizzı, Tahdhıb al-kamal,vol. 33, pp. 370–376; vol. 10, pp. 85–89; for Maslama, see Ibn H. ajar, Lisan al-mızan,vol. 6, p. 715; and for al-Waqidı, see above, note 63.


required of the Christians in developing it amounted at most to connect-ing the available motifs to each other. Even this might not have beennecessary; countless h. adıths, once well known, must be lost today, andamong them might have been a story more similar to the Christian leg-end. Indeed, most of the relevant Islamic traditions are poorly attested.Were it not for the extensive collection of h. adıths about Muh.ammad’sdeath in Ibn Sa↪d’s al-T. abaqat al-kubra, most traditions describing itswidespread denial in the Muslim community and the state of the corpsewhen committed to earth would not have survived.

Not only were the motifs used in the legend present in the Islamictradition, but they were available at the right time and place. While theyawma ’l-arbi ↪a↩ traditions and the h. adıths about the early Muslims’widespread reluctance to admit Muh. ammad’s death apparently circu-lated in the entire Caliphate from the early eighth century onwards,80

the h. adıths about the decaying corpse of Muh. ammad at the time of theburial are attested by the turn of the seventh and eighth century in Kufaand Bas.ra,81 and the ascension traditions by the middle of the eighthcentury only in Kufa.82 In view of the small number of extant traditionswe cannot exclude the possibility that the latter traditions circulatedearlier and elsewhere too. It is significant, however, that they did cir-culate at the same place, in Iraq, where most versions of the Christianlegend were recorded, and at the time, during the early eighth century,to which its earliest traceable version can be dated.

The oldest traces of the Christian legend of Muh. ammad’s death andof the Muslim narrative from which they borrowed are, then, mostlydatable to the first half of the eighth century, and they all vanish fromour sight as we push back into the seventh century. This makes it allthe more intriguing to consider how old the story might be, or, to putit differently, to what extent it might reflect the events that took placein Medina when Muh.ammad died. It is not uncommon among scholarsto regard unorthodox traits of Muh.ammad’s image in the Islamic tradi-tion as going back to the early seventh century in cases when they canbe contrasted with later features that became the classical Islamic po-sition.83 This method, however, does not seem easily applicable to thereconstruction of a sequence of events; after all, not all of their episodesbecame matters of doctrine and controversy in the Muslim community.The only elements of the narratives discussed above that are likely to

80See above, pp. 146–147, 150–152.81See above, pp. 147–149.82See above, pp. 139–146.83See, for example, Friedmann, “Finality of prophethood,” pp. 177–215; and Gold-

feld, “The illiterate Prophet,” pp. 58–67.


reflect historical reality to some extent are therefore those that becamecontentious issues: the belated burial of Muh.ammad, the disagreementsit aroused in the Muslim community, and perhaps ↪Umar’s role in thecontroversy.

As would be expected in hot climate, in the Arabian Peninsula it wasboth customary and necessary to inter the dead soon after their death.Although the date of Muh.ammad’s death varies in the Muslim sources,it may be worth noting that all the dates I have encountered fall in May,June or July. The 24-hour average temperature in Medina during thesemonths is ca. 32–36 C (90–96 F).84 The burials of some early Muslimsare said to have taken place shortly after their death. Abu Bakr, forexample, died on a Tuesday night (or evening) about two years afterMuh.ammad and was interred on the same night.85 Also, al-Suyut.ı (d.911/1505) refers to the late burial of Muh.ammad as one of his specialcharacteristics, so the delay was regarded as exceptional in the Islamictradition itself.86 In view of the general implausibility of anyone in theArabian Peninsula not being buried on the day of his death or at thelatest on the day following it, it is improbable that traditions aboutMuh.ammad’s late burial were invented. Also, it can be easily imaginedthat before somebody as important as Muh.ammad was interred, hisfollowers wanted to ascertain that he had indeed passed away, and beforemodern medical facilities were available, only signs of decompositioncould serve as infallible evidence of death. Muslims, in Iraq or elsewhere,had no reason to invent such traditions. No party in early Islam couldhave gained anything by inventing the delay in the burial of Muhammadin order to blame some of the protagonists: one way or another, theheroes of all of them were involved in it. Once they were in circulation,however, some traditionists could as well continue to transmit them.After all, nobody doubted that Muh.ammad was a mortal,87 and shouldaccordingly have died as one. The yawma ’l-thulatha↩ tradition mighthave been put into circulation in order to contest this version of events.

84See www.worldclimate.com (Madinah, Saudi Arabia). The average temperatureduring daytime would naturally be much higher than this. The data was collected be-tween 1956 and 1990. It should be mentioned, however, that the average temperaturecould have been different in the seventh century CE.

85See the traditions quoted, for instance, in T. abarı, Ta↩rıkh, vol. 1/4, p. 2130, andin Ibn ↪Asakir, Ta↩rıkh madınat Dimashq, vol. 30, pp. 431–434. Another example forthe custom of quick burial is that of Sukayna bint al-H. usayn b. ↪Alı who also died inMedina, in 117 AH (736 CE). She died on a summer day, and since her funeral hadto be postponed until the evening (or night, depending on the variant), frankincense(t.ıb, bukhur, a↪wad) was burnt by her corpse to dispel the smell (or fearing that thecorpse might smell) (see Baladhurı, Ansab al-ashraf [1], vol. 2, pp. 141–142).

86Suyut.ı, Khas. a↩is. , vol. 2, p. 485.87Cf. Qur↩an 3:144 and 18:110.


That a relatively new community would split after the death of itsleader is not surprising. The array of the versions of the ↪Umar story canbe plausibly explained with an underlying attempt to trivialize a splitthat Muh.ammad’s death caused, and that might have involved the en-tire early Muslim community, including its leaders. The most widespreadversion of the ↪Umar story probably developed as a result of ascribing thebelief of many to one, thus representing the division as inconsequentialand saving other Companions from what soon proved to be a discom-fortingly mistaken view. In addition to the split of the community itself,↪Umar’s association with the wrong party was likely not invented either.I am not aware of any group in the early eighth century, when versions ofthe ↪Umar story already circulated, that was interested in presenting AbuBakr or ↪Abbas as superior to ↪Umar. The events, therefore, might bereconstructed as follows. The Muslim community, including its leaders,split as a result of Muh.ammad’s death, with the more powerful group,led by ↪Umar, denying that Muh.ammad’s death was irreversible, andopposing his burial. Due to their influence, Muh.ammad’s interment waspostponed. As nature started taking its course on the corpse, however,the position of this group was exposed as fallacious.88

Whether the events took place as suggested above cannot be proven.But regardless of what really happened in Medina when Muh.ammaddied, the Christian legend largely agreed with what many Muslims inthe eighth century themselves told about Muh.ammad’s death. It wasnot the contents that made the story polemical, but the difference be-tween the religious worldview of its first, Muslim audience and that ofthe new, Christian one. Early Muslims saw Muh.ammad as a mere mortalwith a divine message, and apart from the misguided judgment of theirleaders after Muh.ammad’s death, they probably did not find anythingembarrassing in his ordinary death. But soon after the dead Prophet’sfollowers conquered the Near East, the predominantly Christian inhabi-tants of these lands compared the story of Muh.ammad’s life and deathto the stories of the life and death of those whom they believed he shouldhave resembled: the biblical prophets, Jesus and Christian saints. Com-paring Muh.ammad’s death to that of the biblical prophets was not detri-mental to his standing; apart from Enoch and Elijah who, according tothe Bible, were translated to heaven alive, the rest were believed to havedied the ordinary death of a mortal. But Muh.ammad, in Christian eyes,did not fare well in comparison with Christ. One of the most signifi-cant contrasts between their stories was the end of their lives: the onedisintegrating in the earth, the other being resurrected from the dead

88For a different reconstruction of the events, see Madelung, Succession, pp. 356–360.


and ascending to heaven. When compared to the saints, Muh.ammadalso failed the test. According to his own followers, Muh.ammad’s bodyputrefied in death; the corpses of the saints, according to their vitae,resisted decay, emitted sweet fragrance, and their complexions remainedfresh. The all too-human death of Muh. ammad, according to the religiousworldview of the Christians, did not fit his claim to divine authority. Theeighth-century Muslim narrative of Muh. ammad’s death, as understoodby Christians, made him look inferior to Jesus and the saints, and wasthus a useful polemical argument.

It is tempting to believe that some Islamic traditions were invented inorder to eliminate direct comparison between the stories of Muh.ammad’sand Christ’s death. According to the Christian legend, the Companions’expectation of Muh.ammad’s resurrection on the third day after his deathwas the reason that they did not bury him. The yawma ’l-thulatha↩

tradition might have been devised to contest the legend’s factual basis.89

Had Muh.ammad been buried yawma ’l-thulatha↩, on the second day, notyawma ’l-arbi ↪a↩, on the third day after his death, nobody would havebelieved the Christians that the Companions waited for Muh.ammad’sresurrection.90

Other traditions that might have been invented in response to theChristian legend adopted a different strategy. They maintained thatMuh.ammad’s resurrection, which the Companions had expected accord-ing to the Christian story, indeed took place. One group of these com-prises the h. adıths about Muh.ammad’s resurrection and subsequent lifein his tomb. According to these traditions, Muh.ammad, instead of ris-ing from the dead while not yet buried, was resurrected later, in hisgrave, and continues to be alive there.91 Moreover, the ascension tradi-tions acutely resonate with Christ’s story. According to some of them,Muh.ammad ascended to heaven three days, according to others, fortydays after his death.92 Three and forty days were not selected simplybecause they are topoi ; in this case we would encounter traditions aboutMuh.ammad’s ascension after seven days too. They were selected in or-der to create a parallelism with the resurrection and ascension of Christas told in the Gospels: Christ rose from the dead on the third day afterhis burial, ascended to heaven from his tomb, then appeared to his dis-ciples, stayed with them for forty days, and ascended to heaven again,

89For the yawma ’l-thulatha↩ tradition, see above, pp. 146–147.90Another reason for the invention of the yawma ’l-thulatha↩ tradition could have

been an attempt to bring Muh. ammad’s burial into harmony with Islamic law. Inthe late eighth century, Islamic law prescribed quick burial; see Halevi, Muh. ammad’sgrave, pp. 158–159.

91See, for example, some of the traditions collected in Bayhaqı, H. ayat al-anbiya↩.92See above, pp. 139–145.


in their presence.93

Traditions countering the claims that Muh.ammad’s corpse putrefiedbefore burial also circulated. A h. adıth quotes ↪Alı exclaiming while wash-ing Muh.ammad’s corpse, “You are dearer to me than my father and mymother! How fragrant you are alive and dead.” The h. adıth ends with thewords, “Nothing was observed on (the corpse of) the Messenger of Godof what is (usually) observed on the dead.”94 Showing its popularity,many variants of this tradition circulated; no biography of Muhammadfails to include at least one of them. The pleasant aroma of the deadbody of the holy man is a common motif in Christian saints’ vitae. Forexample, Antonius, the biographer of Simeon Stylites (d. 459), writesabout the holy man’s corpse, “throughout his body and his garmentswas a scented perfume which, from its sweet smell, made one’s heartmerry.”95

About a century ago, Carl H. Becker suggested that Christian pole-mic against Islam in the eighth century influenced the formation of Is-lamic theology.96 More recently, Sarah Stroumsa proposed that the genreof dala↩il al-nubuwwa developed in response to the non-Muslim commu-nities’ probing into Muh.ammad’s prophethood.97 A comparison of thevarious ways the Muslim community remembered Muh.ammad’s deathalso reveals an impact of Christian polemic on Islam. As we saw above,Christian polemic, and the Christian legend of Muh.ammad’s death it-self, might have influenced the formation of Islamic narratives aboutMuh.ammad’s death. Although none of the aforementioned Islamic tra-ditions contains any hint to the Christian legend (or to Christian polemic,

93So at least in the writings of Luke; see Luke 9:22, 18:33, 24:7, 21, 46; Acts 1:3,9–12, 10:40. Only the Book of Acts refers to Jesus’ ascension to heaven after fortydays. The resurrection on the third day is, of course, mentioned also elsewhere inthe New Testament; see, for example, Matthew 16:21, 17:23, 20:19, 27:64; Mark 9:31,10:34, and 1 Corinthians 15:4.

94Ibn Hisham, al-Sıra al-nabawiyya, vol. 1/2, pp. 1018–1019; translations fromGuillaume, The life of Muh. ammad, p. 688 (modified) (bi-abı anta wa-ummı maat.yabaka h. ayyan wa-mayyitan, and wa-lam yura min rasuli Allahi shay↩un mimmayura mina al-mayyiti). These are the last sentences of a h. adıth about the washing ofMuh. ammad’s corpse. Most of the other passages quoted in this paragraph are alsoparts of longer traditions. See also T. abarı, Ta↩rıkh, vol. 1/4, p. 1831; Ibn al-Jawzı,Muntaz.am, vol. 2, p. 479.

95Doran, Lives, p. 98. A common Islamic tradition, “God has forbidden the earthto eat away the bodies of the prophets (inna Allaha [qad ] h. arrama ↪ala ’l-ard. i anta↩kula ajsada al-anbiya↩i)” appears in several h. adıth collections; see, for example,Ibn Maja, Sunan, vol. 1, pp. 524, 345; Nasa↩ı, Sunan, vol. 3, pp. 63–64; and Darimı,Sunan, vol. 2, p. 981. Probably this too developed under the influence of Christiansaints’ lives.

96Becker, “Christian polemic,” pp. 242–257.97Stroumsa, “The signs of prophecy,” pp. 105–106, 114.


even to Christians), it is suggestive that they all circulated at the sametime and the same place. While intra-Muslim debate, triggered by con-verts from Christianity who probably imported their religious worldviewinto Islam, did certainly contribute to the development of the foregoingtraditions as well, the concentration of the relevant traditions in southernIraq (and not in the H. ijaz where a similarly great proportion of Islamictraditions were recorded) points to polemic as the more important fac-tor. The polemical milieu of southern Iraq was more likely to createpolarized opinions than the more uniform society of the H. ijaz.

The Christians strove to preserve the account of Muh. ammad’s or-dinary death just as eagerly as the Muslims wanted to forget it. Thevarious versions of the Christian legend in fact imply more contact withthe eighth-century Islamic tradition than they would seem at first sight.It seems that they did not develop from a single original Christian leg-end; some of them are related to different h. adıths independently fromthe others. The story of Muh.ammad’s death in the Apology of al-Kindıis the richest of them. It gives four different accounts, probably all basedon Islamic traditions.98 According to the Qashun document, the Istoriaand the Adnotatio, the “disciples” were guarding Muh.ammad’s corpse tosee what happens to it. This echoes the h. adıth about the Companionsdeciding to lie in wait (tarabbas. u) to see what happens to the body.99

The different versions of the Christian legend locate the events in twodifferent places; some in a closed room, others in Muh.ammad’s gar-den.100 The former parallels the usual scene of Muh.ammad’s death andfuneral in the Islamic tradition, the room of ↪A↩isha, “The Messengerof God was in his house, his matter is not completed yet, his familyclosed the door on him,” as Ibn Hisham tells us, in the first episode ofyawm al-saqıfa.101 The latter echoes a rare tradition according to whichthe Muslim community prayed over Muh.ammad’s body in “the garden”(bi-wasat.i al-rawd. ati).102

98In addition to the one discussed above, pp. 132–134, 137–138 (unattributed), hequotes three others ascribed to various transmitters; one is attributed to Abu Najıd↪Imran b. (al-)H. us.ayn al-Khuza↩ı (Companion, later qad. ı in Bas.ra, d. 52/672–3),another to a certain D. amran (perhaps a corruption of Shuqran, Muh. ammad’s freedslave, n.d.), still another to “one of them” (see Tartar, Dialogue islamo-chretien,pp. 92–93; and Tien, Risala, pp. 109–110). About ↪Imran and Shuqran, see Mizzı,Tahdhıb al-kamal, vol. 22, pp. 319–321; and vol. 12, pp. 544–546.

99See above, p. 136, and p. 148, note 65.100See above, p. 135.101Ibn Hisham, al-Sıra al-nabawiyya, vol. 1/2, p. 1013 (wa-rasulu Allahi fı baytihi

lam yufragh min amrihi qad aghlaqa dunahu al-baba ahluhu).102Samhudı, Khulas.at al-wafa↩, vol. 1, p. 236. Al-Samhudı does not name his source,

and I was unable to trace it. Elsewhere the Muslims pray over the body in ↪A↩isha’sroom where Muh. ammad died and was buried.


To conclude, the Christian legend exhibits intimate knowledge of theIslamic tradition, not the ignorance of it, as is usually supposed. TheChristians, instead of inventing the story, borrowed it from the Muslimsin the eighth century, and have tenaciously preserved it despite subse-quent changes in the Islamic narrative of Muh.ammad’s death. Instead ofmisrepresenting Islam at will or out of ignorance, the Christians selectedfrom the wealth of h. adıths those that made it the least desirable for theircoreligionists to convert to Islam. Of course, they did tell stories with nohistorical basis about Islam. One can hardly imagine, for example, anyfoundation for the episode appended to some versions of the legend inwhich dogs devour Muh.ammad’s corpse. But such inventiveness is rarerthan it appears at first sight. It was not necessary. There were enoughdifferences between the religious worldviews of Christianity and Islam,between their concepts of sanctity, to make some Islamic narratives func-tion as polemical stories for Christians without modifying them. Whatthe one saw as praiseworthy, the other regarded as despicable. What theone held acceptable, the other thought of as shameful. The story aboutMuh.ammad’s ordinary mortality is an example of such a narrative.

Appendix: The Qashun document

The Qashun document, a fascinating diatribe of an anonymous Chris-tian author against Islam, deserves more attention than it has so far re-ceived in the study of Christian-Muslim relations.103 Its most prominentthemes are episodes from Muhammad’s life, such as stories of a Chris-tian and a Jew instructing him, his miracles, his death and his forgingof a scripture. It also includes an unusually long account of the h. ajjand shorter ones of Muslim prayer and of Muslim attitude to Christ andChristianity. The author attributes this material to a Muslim convert toChristianity.104 Indeed, the text gives some astonishingly detailed andverifiably accurate descriptions of Muslim rituals which distinguishes itfrom most Christian writings on Islam and makes the contribution of aneyewitness conceivable. At the same time, it includes Christian narra-tives unattested in Muslim tradition. Some of these, such as the story of103For the English translation of the text, see Thomson, “Muh. ammad,” pp. 846–853

(the main text is the translation of a later adaptation of the document; the footnotesgive the variants of the Qashun document). Thomson calls the text “Karshuni doc-ument” which is probably incorrect (see below). Since I do not know Armenian, Ifully rely on this English version. Macler, “L’Islam dans la litterature armenienne,”discusses the text briefly; I am not aware of other studies.104See Thomson, “Muh. ammad,” p. 846, n. a, p. 849, n. w., p. 853.


Muhammad’s instruction by a Christian monk, are widespread in Chris-tian polemical literature, while others, like the description of the Ka↪baas a center of snake cult, are entirely unknown.

This work today survives only in a medieval Armenian translation.The date of rendering is unknown, but its terminus ante quem can be setto the twelfth century: although the oldest extant manuscript dates onlyfrom 1273, Mxit’ar of Ani, an Armenian historian who wrote at the turnof the twelfth and thirteenth centuries, already used the document ashis main source on the rise of Islam probably in the same translation.105

Despite this relatively late and solely Armenian attestation, I wouldlike to suggest here that the Qashun document (or a substantial sourceof it) was originally written not on the fringes but in the center of theIslamic world, most probably in Iraq, in Arabic or in Syriac, and no laterthan the early ninth century. The main reason to think so is that thetext contains many motifs unattested in Armenian texts on Islam, andseveral of these occur in Christian Arabic and Syriac writings produced inIraq in the early ninth century. Furthermore, some details may indicateanother source written much earlier by a Melkite monk in the Sinai.Disentangling the layers of the text, however, would require detailedcomparison with a wide range of Muslim and Christian writings by anArmenian specialist. What follow are an amateur’s tentative suggestions,an attempt to draw attention to the value of this document.

Five episodes recall Iraqi Christian polemic against Islam: the sto-ries of Muhammad’s instruction by a Christian monk, then by a Jew, hisdeath, his forging of a holy scripture, and his enjoining of the Muslimsto pray seven times a day.106 All these five episodes occur in the earliestSyriac version of the Christian Bah. ıra legend, and three of them alsoin the Apology of al-Kindı. The details are somewhat different in thevarious versions, but when datable, the Qashun document contains anearlier variant. For example, its author calls the Christian monk Sar-gis. This is the monk’s name in the Apology and in the ninth-centuryWest-Syrian version of the Christian Bah. ıra legend; the name Bah. ırafor Muh.ammad’s Christian teacher appears only in the following cen-tury. Also, the Qashun document does not connect these episodes intoa continuous narrative as the Christian Bah. ıra legend does, and thusprobably reflects an earlier stage in its development. Since the earliestversion of the Christian Bah. ıra legend was in all likelihood written inthe 810s and the Apology of al-Kindı was probably written in the 820s,

105On the manuscript of the Armenian translation and its use by Mxit’ar, see ibid.,pp. 844–845.106For the first and the second episodes, see ibid., p. 846 (cf. n. c), for the third, p.

850, for the fourth, pp. 852–853, and for the fifth, p. 853, n. tt.


both in Iraq,107 the Qashun document or the source its author used forthese episodes probably antedates it and originated in the same region.

The author’s description of his source also seems to point to the earlyninth century as the terminus ante quem for either the composition ofthe document or for one of its major sources, “All this one of Mahmed’sdisciples revealed to us, who had been himself an eye-witness of it all.And terrified by the appearance of the demons, he fled to the island ofCrete; and there he became a Christian and believed in Christ.”108 Asremarked above, it is conceivable that a Muslim convert to Christianityis responsible for the detailed information on Muslim rituals, and theidiosyncrasy of the reference to Crete as the site of the conversion makesthe note sound authentic. Crete was occupied by Andalusian Arabs in827, and it remained under Islamic rule until 961;109 in the interveningperiod no one would have fled there to convert to Christianity. Sincethe contents of the document make it unlikely that it was written in thesecond half of the tenth century or later, it must predate 827.

Other details perhaps point to another, earlier source from the Sinai.Apart from people and locations from Muhammad’s biography and theBible, the geographical focus of a part of the Qashun document is inthe Eastern Mediterranean: it mentions the Mt. Sinai, Egypt, Alexan-dria, the Egyptians, and the Damascenes.110 For instance, accordingto the text, Abraha’s campaign against Mecca (referred to as war be-tween the Ethiopians and the Arabs of Mecca) is related “in Egypt bytradition down to today.”111 But since the campaign was a standardpart of Muhammad’s biography and as such it was related everywherein the Islamic world, this remark makes better sense if read as com-ing from somebody who knew about it only as an Egyptian tradition,i.e., who spent his life in Egypt or nearby and had no opportunity tobecome acquainted with Muslim traditions elsewhere. The note thatMuhammad came to Mt. Sinai112 is best understood from the perspec-tive of an author writing on Mt. Sinai or nearby. The author writesthat Sergius, the monk who instructed Muhammad, was an Arian.113

The Arianism of Muhammad’s teacher is first attested among SyrianMelkites in the early eighth century, but remains unknown among theChristians of Iraq, even later. Also, the text presents Islam in general

107See above, p. 132.108Thomson, “Muh. ammad,” p. 849, n. w; for two other similar notes, see ibid., p.

846, n. a, and p. 853.109See “Ik. rit.ısh,” EI 2, s.v.110Thomson, “Muh. ammad,” pp. 846, 848.111Ibid., p. 848 (italics mine).112Ibid., p. 846.113Ibid., p. 846, n. c.


and the h. ajj in particular as demon worship.114 Three Christian writ-ings from late seventh-century Syria refer to the Muslims as companionsof demons: two Syriac ones written probably in Syria and one Greek textfrom the Sinai.115 Furthermore, the document refers to “Yathrib Med-ina” as the capital of the Arabs.116 Since after the mid-seventh centuryMedina never again became a noteworthy political center, this might bea residue of a seventh century source. Taken together, these details maypoint to a seventh-century Melkite source from the Sinai which couldhave been used by the ninth century Iraqi author whose work was inturn translated into Armenian some time prior to the end of the twelfthcentury.

Finally, it should be noted that Thomson’s dubbing the text “Kar-shuni document” is unlikely to be correct. With this appellation, hefollowed Macler who, reviewing the edition of the text and noting thatthe Armenian version refers to it as “extrait de Qachoun” or “recueil deQachoun” (i qachounen qaghadzou), suggested that Qashun should betaken as a corruption of the word Karshuni.117 But this is unlikely, fortwo reasons. First, while dated Karshuni glosses are already attested inthe twelfth century (only a few undated ones could be earlier), the oldestmanuscripts containing texts copied in their entirety in Karshuni datefrom as late as the thirteenth century, after the earliest attestation ofthe Qashun document in Armenian translation. Second and even moreimportantly, the name Karshuni first appears much later, in the sixteenthcentury, and then in the form “Garshuni.”118 It is thus improbable thatthe Armenian translation, dating from the twelfth century at the latest,was made from a Karshuni text and referred to itself as Karshuni. ASyriac or Arabic Vorlage is much more likely.

114Ibid., pp. 846, 849.115Reinink, “The doctrine of the demons,” pp. 132–134.116Thomson, “Muh. ammad,” p. 847.117Macler, “L’Islam dans la litterature armenienne,” pp. 494 and 522.118This discussion owes much to the papers given in the session on Karshuni at the

Eighth Conference of Christian Arabic Studies, held in Granada (Spain), September26–27, 2008: Emanuela Braida, “Garshuni manuscripts and Garshuni notes in Syriacmanuscripts,” Gregory Kessel, “The importance of the manuscript tradition of the‘Book of Grace’ (7th c.) for the study of Garshuni,” and Ray Mouawad “Maronitesand the Garshuni script.”


Bibliography

↪Abd al-Razzaq, Mus.annaf =↪Abd al-Razzaq al-S. an↪anı. Al-Mus.annaf. Beirut, 1421/2000.

Abu Ja↪far al-Qummı, Bas. a↩ir al-darajat =Abu Ja↪far al-Qummı. Bas. a↩ir al-darajat fı fad. a↩il al Muh. ammad.Qom, 1426/2005.

Abu Nu↪aym, Dala↩il al-nubuwwa =Abu Nu↪aym al-Is.bahanı. Dala↩il al-nubuwwa. H. aydarabad, 1369/1950.

——, H. ilyat al-awliya↩ =idem. H. ilyat al-awliya↩ wa-t.abaqat al-as.fiya↩. Cairo, 1357/1938.

↪Ajlunı, Kashf al-khafa↩ =Al-↪Ajlunı. Kashf al-khafa↩ wa-muzıl al-ilbas ↪amma ’shtahara minal-ah. adıth ↪ala alsinat al-nas. Aleppo, n.d.

Baladhurı, Ansab al-ashraf (1) =Al-Baladhurı. Ansab al-ashraf. M. H. amıd Allah, ed. Cairo, 1987.

——, Ansab al-ashraf (2) =idem. Ansab al-ashraf. Damascus, 1996.

Bayhaqı, H. ayat al-anbiya↩ (1) =Al-Bayhaqı. Kitab ma warada fı h. ayat al-anbiya↩ ba↪d wafatihim.Beirut, 1410/1990.

——, H. ayat al-anbiya↩ (2) =idem. H. ayat al-anbiya↩ s.alawat Allah ↪alayhim ba↪da wafatihim.Medina, 1414/1993.

Becker, “Christian polemic” =Becker, C.H. “Christian polemic and the formation of Islamic dog-ma.” In R. Hoyland, ed. Muslims and others in early Islamicsociety. Aldershot, 2004, pp. 241–257.

Colbert, The martyrs of Cordoba =Colbert, E.P. The martyrs of Cordoba (850–859): a study of thesources. Washington D.C., 1962.

Daniel, Arabs and mediaeval Europe =Daniel, N. Arabs and mediaeval Europe. London, 1975.


——, Islam and the West =idem. Islam and the West: the making of an image. Oxford, 1993.

Darimı, Sunan =Musnad al-Darimı al-ma↪ruf bi-Sunan al-Darimı. Riyad. , 1421/2000.

Daylamı, Firdaws =Al-Daylamı. al-Firdaws bi-ma↩thur al-khit.ab. Beirut, 1406/1986.

Dıaz y Dıaz, “Los textos antimahometanos” =Dıaz y Dıaz, M.C. “Los textos antimahometanos mas antiguos encodices espanoles.” Archives d’histoire doctrinal et litteraire duMoyen Age 37 (1970): 149–168.

Diyarbakrı, Ta↩rıkh al-khamıs =Al-Diyarbakrı. Ta↩rıkh al-khamıs fı ah. wal anfas nafıs. Beirut,1970.

Doran, Lives =Doran, R. tr. The lives of Simeon Stylites. Kalamazoo, Mich.1992.

Fasawı, al-Ma↪rifa wa-’l-ta↩rıkh =Al-Fasawı. Kitab al-ma↪rifa wa-’l-ta↩rıkh. Baghdad, 1394/1974.

Franke, Die freiwilligen Martyrer =Franke, R.F. Die freiwilligen Martyrer von Cordova und das Ver-haltnis der Mozaraber zum Islam (nach den Schriften des Sperain-deo, Eulogius und Alvar). Munster, 1958.

Friedmann, “Finality of prophethood” =Friedmann, Y. “Finality of prophethood in Sunnı Islam.” JSAI 7(1986): 177–215.

Ghazalı, al-Durra al-fakhira =Al-Ghazalı. Al-Durra al-fakhira fı kashf ↪ulum al-akhira. Beirut,1407/1987.

Gil, Corpus scriptorum muzarabicorum =Gil, J. Corpus scriptorum muzarabicorum. Madrid, 1973.

Goldfeld, “The illiterate Prophet” =Goldfeld, I. “The illiterate Prophet (nabı ummı): an inquiry intothe development of a dogma in Islamic tradition.” Der Islam 57(1980): 58–67.


Gonzalez Munoz, Exposicion y refutacion del Islam =Gonzalez Munoz, F. Exposicion y refutacion del Islam: la versionlatina de las epıstolas de al-Hasimı y al-Kindı. A Coruna, 2005.

Guillaume, The life of Muh. ammad =Guillaume, A. The life of Muh. ammad: a translation of [Ibn] Ish. aq’sSırat rasul Allah. Karachi, 1997.

Halevi, Muh. ammad’s grave =Halevi, L. Muh. ammad’s grave: death rites and the making of Is-lamic society. New York, 2007.

H. ashiyat al-Dasuqı =H. ashiyat al-Dasuqı ↪ala ’l-sharh. al-kabır. Cairo, n.d.

Hoyland, Seeing Islam =Hoyland, R.G. Seeing Islam as others saw it: a survey and evalua-tion of Christian, Jewish and Zoroastrian writings on early Islam.Princeton, 1997.

Ibn ↪Abd al-Hadı, al-S. arim al-munkı =Ibn ↪Abd al-Hadı. Al-S. arim al-munkı fı ’l-radd ↪ala ’l-Subkı. Bei-rut, 1412/1992.

Ibn Abı l-H. adıd, Sharh. =Ibn Abı l-H. adıd. Sharh. Nahj al-balagha. Qom, 1404/1985.

Ibn Abı Shayba, Mus.annaf =Ibn Abı Shayba. Al-Kitab al-mus.annaf fı ’l-ah. adıth wa-’l-athar.Riyad. , 1409/1988.

Ibn ↪Asakir, Ta↩rıkh madınat Dimashq =Ibn ↪Asakir. Ta↩rıkh madınat Dimashq wa-dhikr fad. liha wa-tasmi-yat man h. allaha min al-amathil aw ijtaza bi-nawah. ıha min wari-dıha wa-ahliha. ↪Alı Shırı, ed. Beirut, 1995–2000.

Ibn Babawayh, Faqıh =Ibn Babawayh. Kitab man la yah. d. uruhu al-faqıh. Qom, 1404/1983.

Ibn H. ajar, Fath. al-barı =Ibn H. ajar al-↪Asqalanı. Fath. al-barı bi-sharh. al-Bukharı. Cairo,1378/1959.

——, Lisan al-mızan =idem. Lisan al-mızan. Beirut, 1416/1995.


——, Raf ↪ al-is.r =idem. Raf ↪ al-is.r ↪an qud. at Mis.r. Cairo, 1418/1998.

——, Talkhıs. al-h. abır =idem. Talkhıs. al-h. abır fı takhrıj ah. adıth al-Rafi ↪ı al-kabır. Beirut,1419/1998.

Ibn H. ibban, Kitab al-majruh. ın =Ibn H. ibban. Kitab al-majruh. ın min al-muh. addithın. Riyad. , 1420/2000.

Ibn Hisham, al-Sıra al-nabawiyya =Das Leben Muhammed’s nach Muhammed Ibn Ishak bearbeitet vonAbd el-Malik Ibn Hischam. F. Wustenfeld, ed. Gottingen, 1858.

Ibn al-Jawzı, Muntaz.am =Ibn al-Jawzı. Al-Muntaz.am. Beirut, 1358/1939.

Ibn Kathır, al-Bidaya wa-’l-nihaya =Ibn Kathır. Al-Bidaya wa-’l-nihaya. Beirut, 1415/1994.

Ibn Khallikan, Wafayat al-a↪yan =Ibn Khallikan. Wafayat al-a↪yan wa-anba↩ abna↩ al-zaman. Bei-rut, 1419/1998.

Ibn Maja, Sunan =Sunan al-H. afiz. Abı ↪Abd Allah Muh. ammad b. Yazıd al-Qazwını IbnMaja. N.p., 1972.

Ibn Manz.ur, Mukhtas.ar ta↩rıkh Dimashq =Ibn Manz.ur. Mukhtas.ar ta↩rıkh Dimashq li-Ibn ↪Asakir. Damas-cus, 1984.

Ibn al-Mulaqqin, al-Badr al-munır =Ibn al-Mulaqqin. Al-Badr al-munır fı takhrıj al-ah. adıth wa-’l-atharal-waqi ↪a fı ’l-Sharh. al-kabır. Riyad. , 1425/2004.

Ibn Quluya, Kamil al-ziyarat =Ibn Quluya. Kamil al-ziyarat. Najaf, 1356/1937.

Ibn Qutayba, Ma↪arif =Ibn Qutayba. Al-Ma↪arif. Cairo, 1969.

Ibn Sa↪d, T. abaqat =Ibn Saad. Biographien Muhammeds, seiner Gefahrten und derspateren Trager des Islams. E. Mittwoch et al., eds. Leiden, 1904–1905.


Ibn al-Zaghunı, Id. ah. =Ibn al-Zaghunı. Al-Id. ah. fı us. ul al-dın. Riyad. , 1424/2003.

Juwaynı, Nihayat al-mat.lab =Al-Juwaynı. Nihayat al-mat.lab fı dirayat al-madhhab. Jeddah,1428/2007.

Karajikı, Kanz al-fawa↩id =Al-Karajikı. Kanz al-fawa↩id. Beirut, 1405/1985.

Kohlberg, “Western accounts” =Kohlberg, E. “Western accounts of the death of the Prophet Mu-h. ammad.” In M.A. Amir-Moezzi and J. Scheid, eds. L’Orient dansl’histoire religieuse de l’Europe: l’invention des origines. Turn-hout, 2000, pp. 165–195.

Koningsveld, “The apology of al-Kindı” =Koningsveld, P.S. van. “The apology of al-Kindı.” In T.L. Het-tema and A. van der Kooij, eds. Religious polemics in context:papers presented to the Second International Conference of the Lei-den Institute for the Study of Religions (LISOR) held at Leiden,27–28 April 2000. Assen, 2004, pp. 69–92.

Kulaynı, Kafı =Al-Kulaynı. Al-Furu↪ min al-kafı. Tehran, 1362–3/1943–4.

Landron, Attitudes nestoriennes =Landron, B. Chretiens et musulmans en Irak: attitudes nestori-ennes vis-a-vis de l’islam. Paris, 1994.

Macler, “L’Islam dans la litterature armenienne” =Macler, F. “L’Islam dans la litterature armenienne: d’apres la pub-lication recente du ‘Livre des questions’ de Tathewatsi.” Revue desetudes islamiques 6 (1932): 493–522.

——, “Un document armenien” =idem. “Un document armenien sur l’assassinat de Mahomet parune juive.” In Melanges Hartwig Derenbourg (1844–1908): recueilde travaux d’erudition dedies a la memoire d’Hartwig Derenbourgpar ses amis et ses eleves. Paris, 1909, pp. 287–295.

Madelung, Succession =Madelung, M. The succession to Muh. ammad: a study of the earlyCaliphate. New York, 1997.


Majlisı, Bih. ar al-anwar =Al-Majlisı. Bih. ar al-anwar. Tehran, 1376/1957.

Mingana (ed.), “Apocalypse of Peter” =Mingana, A., ed. “Apocalypse of Peter.” Woodbrooke Studies 3(1931): 93–449.

Mizzı, Tahdhıb al-kamal =Al-Mizzı. Tahdhıb al-kamal fı asma↩ al-rijal. Bashshar ↪AwwadMa↪ruf, ed. Beirut, 1992.

Najashı, Rijal =Rijal al-Najashı: ah. ad al-us. ul al-rijaliyya. Beirut, 1408/1988.

Nasa↩ı, Sunan =Sunan al-Imam al-H. afiz. ↪Abd al-Rah. man Ah. mad b. Shu↪ayb b. ↪Alıal-Khurasanı al-Nasa ↩ı. Beirut, 1416/1995.

(Al-)Qad. ı ↪Iyad. , Shifa↩ =Al-Qad. ı ↪Iyad. . al-Shifa bi-ta↪rıf h. uquq al-Mus.t.afa. Cairo, 1977.

Qarafı, Dhakhıra =Al-Qarafı. Al-Dhakhıra. Beirut, 1994.

Rafi↪ı, al-Sharh. al-kabır =Al-Rafi↪ı. Al-↪Azız sharh. al-wajız al-ma↪ruf bi-’l-sharh. al-kabır.Beirut, 1417/1997.

Reinink, “The doctrine of the demons” =Reinink, G.J. “Following the doctrine of the demons: early Chris-tian fear of conversion to Islam.” In J.N. Bremmer, W.J. vanBekkum, and A.L. Molendijk, eds. Cultures of conversions. Leu-ven, 2006, pp. 127–138.

Roggema, “Biblical exegesis” =Roggema, B. “Biblical exegesis and interreligious polemics in theArabic Apocalypse of Peter — The Book of the Rolls.” In D.Thomas, ed. The Bible in Arab Christianity. Leiden, 2007, pp.131–150.

——, The legend of Sergius Bah. ıra =idem. The legend of Sergius Bah. ıra: Eastern Christian apologeticsand apocalyptic in response to Islam. Leiden, 2009.

Safran, “Identity and differentiation” =Safran, J.M. “Identity and differentiation in ninth-century al-An-dalus.” Speculum 76 (2001): 573–598.


S. aghanı, Mawd. u↪at =Mawd. u↪at al-S. aghanı. Damascus—Beirut, 1405/1985.

S. alih. ı, Subul al-huda =Al-S. alih. ı. Subul al-huda wa-’l-rashad fı sırat khayr al-↪ibad. Beirut,1414/1993.

Samhudı, Khulas.at al-wafa =Al-Samhudı. Khulas.at al-wafa bi-akhbar dar al-Mus.t.afa. Cairo,1427/2006.

Samir, “Apologie d’al-Kindı” =Samir, S.K. “La version latine de l’Apologie d’al-Kindı (vers 830ap. J.-C.) et son original arabe.” In C. Aillet, M. Penelas and Ph.Roisse, eds. ¿‘Existe una identidad mozarabe? Historia, lenguay cultura de los cristianos de al-Andalus (siglos IX–XII). Madrid,2008, pp. 33–81.

Shawkanı, Nayl al-awt.ar =Al-Shawkanı. Nayl al-awt.ar min ah. adıth sayyid al-akhyar: Sharh.Muntaqa al-akhbar. Beirut, 1973.

Stroumsa, “The signs of prophecy” =Stroumsa, S. “The signs of prophecy: the emergence and early de-velopment of a theme in Arabic theological thought.” The HarvardTheological Review 78 (1985): 101–114.

Suyut.ı, Khas. a↩is. =Al-Suyut.ı. Al-Khas. a↩is. al-kubra. Beirut, 1405/1985.

——, al-La↩alı al-mas.nu↪a =idem. Al-La↩alı al-mas.nu↪a fı ’l-ah. adıth al-mawd. u↪a. Cairo, n.d.

Szilagyi, “Muh.ammad and the monk” =Szilagyi, K. “Muh.ammad and the monk: the making of the Chris-tian Bah. ıra legend.” JSAI 34 (2008): 169–214 .

T. abarı, Ta↩rıkh =Annales quos scripsit Abu Djafar Mohammed ibn Djarir at-Tabari.M.J. de Goeje, ed. Leiden, 1901.

——, Tafsır =idem. Jami ↪ al-bayan ↪an ta↩wıl ay al-Qur ↩an. Beirut, 1405/1984.

——, The history of al-T. abarı =idem. The history of al-T. abarı: vol. 9, The last years of theProphet. I.K. Poonawala, tr. Albany, 1990.


Tartar, Dialogue islamo-chretien =Tartar, G. Dialogue islamo-chretien sous le calife al-Ma’mun (813-834): les epıtres d’al-Hasimı et d’al-Kindı. PhD. dissertation, Uni-versity of Strasbourg, 1977.

Tha↪labı, Tafsır =Al-Tha↪labı. Al-Kashf wa-’l-bayan (Tafsır al-Tha↪labı). Abu Mu-h. ammad b. ↪Ashur, ed. Beirut, 1422/2002.

Thomson, “Muh.ammad” =Thomson, R.W. “Muh.ammad and the origin of Islam in Armenianliterary tradition.” In D. Kouymjian ed. Armenian studies inmemoriam Haıg Berberian. Lisbon, 1986, pp. 829–858.

Tien, Risala =Risalat ↪Abd Allah ibn Isma ↪ıl al-Hashimı ila ↪Abd al-Masıh. ibnIsh. aq al-Kindı yad ↪uhu biha ila l-islam wa-risalat ↪Abd al-Masıh.ila ’l-Hashimı yaruddu biha ↪alayhi wa-yad ↪uhu ila l-nas.raniyya.A. Tien, ed. London, 1885.

Tolan, Saracens =Tolan, J.V. Saracens: Islam in the medieval European imagination.New York, 2002.

——, “Un cadavre mutile” =idem. “Un cadavre mutile: le dechirement polemique de Ma-homet.” Le Moyen Age: Revue d’histoire et de philologie 104/1(1998): 53–72.

——, “A mangled corpse” =idem. “A mangled corpse: the polemical dismemberment of Mu-h. ammad.” In idem. Sons of Ishmael: Muslims through Europeaneyes in the Middle Ages. Gainesville, 2008, pp. 19–34.

T. usı, Rijal =Rijal al-T. usı. Qom, 1415/1995.

——, Tahdhıb al-ah. kam =Al-T. usı. Tahdhıb al-ah. kam. Teheran, 1417/1997.

Vajda, “Un vestige oriental” =Vajda, G. “Un vestige oriental de “l’anti-biographie” de Moh.am-med et un echo de la tragedie de Karbala.” Revue de l’histoire desreligions 189/2 (1976): 177–180.


Wansharısı, Mi ↪yar =Al-Wansharısı. Al-Mi ↪yar al-mu↪rib wa-’l-jami ↪ al-mughrib ↪anfatawa ahl Ifrıqiya wa-’l-Andalus wa-’l-Maghrib. Rabat, 1401/1981.

Wolf, “The earliest Latin lives” =Wolf, K.B. “The earliest Latin lives of Muh.ammad.” In M. Gerversand R.J. Bikhazi, eds. Conversion and continuity: indigenousChristian communities in Islamic lands, eighth to eighteenth cen-turies. Toronto, 1990, pp. 89–101.

Documents

JERUSALEM STUDIES IN ARABIC AND ISLAM 36(2009)¡gyi, K - A prophet like Jesus.pdf132 Krisztina Szil agyi Muh.ammad’s death according to the Christians ... a Christian Arab belonging